Rebbeca M. Duar

Host-adapted lactobacilli: evolution, molecular mechanisms and functional

applications

Rebbeca M. Duar

A thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Food Science and Technology

Department of Agricultural, Food and Nutritional Science

University of Alberta

Bacteria of the genus Lactobacillus can be found associated with plants, insects and vertebrate hosts, and their lifestyle can range from free-living to strictly host specific. Of the lactobacilli associated with vertebrates, the lifestyle of L. reuteri is particularly well understood.

The species has been studied by population genetics, comparative genomic and functional analyses in animal models. The phylogenetic structure of L. reuteri suggests that lineages evolved alongside with rodents, poultry, swine and humans. For rodent strains, co-evolution resulted in host-adaptation. The first goal of this dissertation was to determine whether host-adaptation extended to non-rodent lineages and also to resolve open questions regarding the evolutionary relationships within lineage VI, which is shared by human and poultry isolates. An experimental approach was devised to determine the ability of strains to propagate under the ecological conditions of the gastrointestinal tract (GIT) of different hosts. Rodent isolates became enriched in the GIT of mice and poultry isolates in chickens. Moreover, human isolates of the lineage VI were found to be competitive in the GIT of chickens but not in humans. These findings revealed that L. reuteri evolved host-specialization in rodents and chicken, while open questions remain about the exact evolutionary consequences in humans and pigs.

Biofilm formation is a common strategy by which lactobacilli maintain stable associations with their hosts. Only rodent isolates of L. reuteri can produce biofilms in the forestomach of mice. The second goal of this dissertation was to determine the role of a rodent-specific two component system (TCS70529-30) in biofilm formation of the rat isolate L. reuteri 100-23. Experiments in monoassociated mice revealed that mutation of the response regulator, but not the histidine kinase impaired biofilm formation. In vitro experiments confirmed in vivo and findings and further revealed significant alterations in the architecture of the mutant biofilms. Compared to the wildtype, histidine kinase mutants produced thick and robust biofilms, while the response regulator mutants formed thinner and less adherent biofilms. These findings provide empirical

ii evidence of rodent specific signal transduction system playing a role in biofilm formation of L reuteri, likely by regulating genes responsible for development of the biofilm matrix.

Contrary to rodent strains, human isolates of L. reuteri lack the genetic machinery to form biofilms, but conserve a 58-gene pdu-cbi-cob-hem cluster (pdu-cluster). Encoded in the pdu- cluster is the PduCDE diol dehydratase involved in utilization of 1,2 propanediol (1,2 PD). In the human gut, 1,2 PD is readily available as a result of fermentation of rhamnose and fucose found in dietary and host-derived glycans, respectively. The third goal of this dissertation was to determine the role of the pdu-cluster in utilization of 1,2 PD by human isolates of L. reuteri. The ability of the human isolate L. reuteri ATCC 6475 to cross-feed from 1,2 PD produced by

Escherichia coli MG1655 and Bifidobacterium breve UCC2003 was determined in vitro and compared to a pduCDE mutant. We found that during fermentation of hexoses, 1,2 PD serves as an electron acceptor increasing the metabolic efficiency of L. reuteri, a factor that could be pivotal to the competitiveness of human isolates of the human GIT.

The fourth goal of this dissertation was to identify and characterize bacterial isolates from the proximal GI tract of pigs capable of degrading peptides involved in the etiology of celiac disease.

Strains were selected from the GIT tract of pigs fed a 20% gluten diet and after an in vitro process aimed to enrich for gluten degrading bacteria. Pigs were selected as these animals harbor large amounts of lactobacilli. Strains of the species L. amylovorus, L. johnsonii, L. ruminis, and L. salivarius were identified as having the highest proteolytic activity against several well characterized gluten immunotoxic peptides. Since these strains are adapted to the conditions in the proximal GI tract, they are likely to be good candidates for probiotics aimed at removing gluten epitopes before they reach the epithelium of the small intestine in celiac patients.

Together findings in this dissertation contribute to our understanding of the evolution of L. reuteri with different vertebrate hosts, reveal insights into lineage-specific functions underlying adaptation to the vertebrate GIT, and provide a basis for the selection of lactobacilli adapted to

GIT for functional applications.

iii

PREFACE

This thesis is an original work by Rebbeca M. Duar

A version of Chapters and 6 – are part of invited review article currenty in revision (April,

2017) to be published in the journal FEMS Microbiology Reviews in a dedicated edition for the

LAB12 conference as: R.M Duar, X B. Lin, J. Zheng, M.E Martino, T Grenier, ME Pérez-Muñoz,

F Leulier, MG Gänzle, J Walter. Lifestyles in transition: Evolution and natural history of the genus

Lactobacillus

R.M.D and X.B.L, M.G.G and J.W contributed equally to this work by conceptualizing the idea, analyzing the data, writing and editing the manuscript. M.G.G and J.Z performed the phylogenomic analyses. M.G.G provided the type strains’ metadata and conducted metagenomics analysis. M.E designed the illustrations M.E.M, T.G and F.L wrote the section on the nomadic lactobacilli and edited the manuscript.

A version of Chapter 2 is publiched as- RM. Duar, SA. Frese, SC. Fernando, TE. Burkey, G

Tasseva, XB Lin. DA. Peterson, J Blom, CQ. Wenzel, CM. Szymanski and J Walter. Experimental evaluation of host adaptation of Lactobacillus reuteri to different vertebrate species Applied and

Environmental Microbiology (2017) 83: e00132-17; doi:10.1128/AEM.00132-17

R.M.D designed the experiments, collected and analyzed the data, and wrote the manuscript.

J.W designed the experiments, supervised data analyses, wrote and edited the manuscript. S.A.F designed the experiments, conducted the mouse experiments and edited the manuscript. S.E.F and T.E.B provided the gnotobiotic pig mode and gave technical advice. CQW and CMS provided the chicken model and gave conceptual and technical advice. D.A.P provided mice and gave technical advice. G.T helped preparing materials and collected data. J.B gave technical support and conceptual advice in for the comparative genomic analyses.

Chapter 3- RM. Duar, XB. Lin, T Grenier, M Bording-Jorgensen, LA. Cody, E Wine, AE.

Ramer-Tait, MG. Gänzle, J Walter. A rodent-strain specific two-component system regulates biofilm formation of Lactobacillus reuteri 100-23. Manuscript in preparation.

R.M.D generated the mutants, designed the experiments, collected and analyzed the data, and wrote the paper. X.B.L provided technical advice during the generation of the mutants and conduced the in vivo experiments. T.G designed and conducted in vitro experiments and performed the SEM sample preparation and image collection.. E.W provided materials for CLSM.

M.B.J provided technical advice for confocal microscopy, collected the images and conduced part of the CLSM image collection. L.A.C conduced the confocal imaging analysis of the in vivo samples. A.E.R provided mice and supervised mouse experiments at the University of Nebraska.

M.G.G provided conceptual and technical advice. J.W conceived the study, conceptualized the experiments and supervised data analysis.

Chapter 4 -RM. Duar, C. Cheng, XB. Lin, S Mohamed, JP van Pijkeren, JH Oh, D Van

Sinderen, MG. Gänzle, J Walter. In vitro cross-feeding of 1,2 propanediol in human isolates of

Lactobacillus reuteri. Manuscript in preparation.

R.M.D conceived the study, designed the experiments, collected and analyzed the data, and wrote the manuscript. C.C and S.M conducted the experiments and collected data. X.B.L provided conceptual advice and performed HPLC analysis. J.P and J.H.O generated the mutant. D.V.S provided the strain Bifidobacterium breve UCC2003 and gave technical and conceptual advice.

MGG and J.W conceptualized the experiments and supervised data analysis.

A version of Chapter 5 is published as- RM. Duar, KJ. Clark, PB. Patil, C. Hernández, S.

Brüning, TE. Burkey, N. Madayiputhiya, SL. Taylor, J. Walter. Identification and characterization

v of intestinal lactobacilli strains capable of degrading immunotoxic peptides present in gluten.

Journal of Applied Microbiology (2015) 118: 515-527; doi:10.1111/jam.12687

R.M.D conducted experiments, collected and analyzed data, and wrote the manuscript. K.J.C,

P.B.P, C.H and S.B conducted experiments, collected and analyzed the data. S.L.T provided materials, conceptual and technical advice during the ELISA analyses. T.E.B provided the pigs and gave technical advice. N.M conduced the mass spectroscopy analysis. J.W conceived the project, conceptualized the experiments, supervised data analysis, wrote and edited the manuscript.

Chapter 6- Conclusions, implications and future directions

ACKNOWLEDGMENTS

I would like to thank my advisor Dr. Jens Walter, for giving me the opportunity to work in his lab, for his time, support and guidance throughout my PhD training. I also want to thank Dr.

Michael Gänzle and Dr. Lynn McMullen for serving in my advisory committee, giving me valuable advice and for providing me with opportunities to teach.

Also, I would like to acknowledge present and past members of the Walter lab especially

Steve and Brooke for their research contributions and for all debates and discussions, both scientific and non-scientific.

During my time in grad school I’ve developed friendships with some truly wonderful individuals, María Isabel, Andrés, Sarah (s), Rae and Mike, thanks for letting me vent, for giving me advise and for making me laugh.

I would like to dedicate my work to my whole family for understanding my dream to study abroad and for their permanent support in my life, especially to my mom and aunt, two strong and independent women I was very fortunate to have as role models. I would like to give a very special thanks to my husband Dallas, for his loving support and for putting up with me during stressful times.

“And, when you want something, all the universe conspires in helping you to achieve it.”

-Paulo Coelho, The Alchemist

vii

TABLE OF CONTENTS

ABSTRACT ...... ii

PREFACE ...... iv

ACKNOWLEDGMENTS ...... vii

TABLE OF CONTENTS ...... viii

LIST OF FIGURES ...... xiv

LIST OF TABLES ...... xvi

GLOSSARY OF TERMS ...... xvii

1. Chapter One: Evolution and lifestyles of species of the genus Lactobacillus ...... 1

1.1 Introduction ...... 1

1.2 Habitats of lactobacilli ...... 2

1.3 What are the lifestyles of lactobacilli in nature? ...... 5

1.3.1 Evolutionary insight through phylogenomics ...... 7

1.3.2 Patterns of genome evolution reflect an evolutionary transition to a symbiotic

lifestyle ...... 9

1.3.3 Metabolic capabilities reflect lifestyle adaptations ...... 10

1.4 Paradigms of Lactobacillus lifestyles ...... 13

1.4.1 Free-living lifestyle ...... 13

1.4.2 Host-adapted lifestyle ...... 14

1.5 A model for the evolution of lifestyle transitions in the Lactobacillus sensu lato ...... 28

1.6 Open questions ...... 31

1.7 Supplementary material ...... 34

1.8 References ...... 39

viii

2. Chapter two: Experimental evaluation of host adaptation of Lactobacillus reuteri to different vertebrate species ...... 53

2.1 Introduction ...... 54

2.2 Materials and Methods ...... 56

2.2.1 Strains, media and growth conditions ...... 56

2.2.2 Preparation of strain mixtures to prepare inocula ...... 56

2.2.3 Mouse experiment ...... 58

2.2.4 Chicken experiment ...... 58

2.2.5 Pig experiment ...... 59

2.2.6 Human subjects ...... 60

2.2.7 Strain typing and identification...... 60

2.2.8 Genome sequencing and annotation ...... 61

2.2.9 Accession numbers ...... 61

2.2.10 Comparative genomic and phylogenetic analyses ...... 61

2.2.11 Ancestral state analysis ...... 62

2.2.12 Statistical analysis ...... 62

2.3 Results ...... 63

2.3.1 Evolutionary relationships of L. reuteri strains using whole genome phylogenetic

analysis ...... 63

2.3.2 Introduction of L. reuteri to the digestive tract of different vertebrate hosts...... 64

2.3.3 Rodent isolates become enriched in the murine host ...... 68

2.3.4 The chicken host ...... 68

2.3.5 The pig host ...... 74

2.3.6 The human host ...... 74

2.3.7 Evolution and genome characteristics of lineage-VI strains ...... 74

2.4 Discussion ...... 79

2.4.1 Lineage VI strains, even if isolated from humans, show elevated fitness in chicken

...... 80

2.4.2 Human and porcine isolates do not show elevated ecological fitness in their

respective hosts ...... 82

2.4.3 Limitations of this study ...... 84

2.5 Conclusion ...... 85

2.6 Acknowledgments ...... 85

2.6.1 Funding information ...... 86

2.7 References ...... 86

3. Chapter three: A rodent-strain specific two-component system is involved in biofilm formation of Lactobacillus reuteri 100-23...... 92

3.1 Introduction ...... 93

3.2 Materials and Methods ...... 95

3.2.1 Bacterial strains, plasmids and media ...... 95

3.2.2 DNA manipulations ...... 96

3.2.3 Generation of L. reuteri 100-23 knockout mutants ...... 96

3.2.4 Mouse experiments ...... 98

3.2.5 In vitro biofilm assays...... 98

3.2.6 Mechanical properties of biofilms ...... 99

3.2.7 Confocal Microscopy...... 99

3.2.8 Scanning Electron Microscopy ...... 100

3.2.9 Statistical analyses ...... 100

3.3 Results ...... 101

3.3.1 Genetic organization of the two component system 70529-30 ...... 101

3.3.2 Deletion of the rr70530 but not hk70529 the resulted in changes in biofilm in vivo

...... 101

3.3.3 Mutation of rr70530 results in biofilm defects in vitro ...... 103

3.3.4 Mutant biofilms exhibit different macroscopic properties and microscale

architectures ...... 104

3.3.5 Matrix architecture is associated with changes in biofilm resistance to sheer stress

...... 107

3.4 Discussion ...... 109

3.5 References ...... 113

4. Chapter four: Metabolic cross-feeding between 1,2 propanediol-producing intestinal bacteria and the human isolate Lactobacillus reuteri ATCC PTA6475 ...... 118

4.1 Introduction ...... 119

4.2 Materials and Methods ...... 121

4.2.1 Bacterial strains and media ...... 121

4.2.2 Generation of the L. reuteri ΔpduCDE mutant ...... 122

4.2.3 Basal media for fermentations ...... 122

4.2.4 Screening of 1.2 PD utilization by L reuteri...... 123

4.2.5 In vitro fermentations for 1,2 PD production ...... 123

4.2.6 Cross-feeding assays ...... 124

4.2.7 Analytical methods ...... 124

4.2.8 Statistical analysis ...... 124

4.3 Results ...... 126

4.3.1 Effect of 1,2 PD on growth of L. reuteri ...... 126

4.3.2 Production of 1,2 PD from deoxy hexoses ...... 127

4.3.3 Cross-feeding of E. coli-produced 1,2 PD ...... 130

4.3.4 Cross-feeding of B. breve-produced 1,2 PD ...... 133

4.4 Discussion ...... 135

4.5 Supplementary material ...... 138

4.6 References ...... 140

5. Chapter five: Identification and characterization of intestinal lactobacilli strains capable of degrading immunotoxic peptides present in gluten...... 146

5.1 Introduction ...... 147

5.2 Materials and Methods ...... 149

5.2.1 Animals and intestinal sampling ...... 149

5.2.2 Enrichment of gluten utilizing bacteria ...... 149

5.2.3 Culture techniques and classification of isolates ...... 150

5.2.4 Initial screen of isolates for specific proteolytic activities ...... 151

5.2.5 Immunotoxic peptides ...... 151

5.2.6 Bacterial strains, culture media, and growth conditions ...... 152

5.2.7 Peptide degradation ...... 152

5.2.8 ELISA tests ...... 153

5.2.9 Mass Spectrometry ...... 153

5.2.10 Statistical analysis ...... 154

5.3 Results ...... 155

5.3.1 Isolation of intestinal bacteria directly from the proximal GI tract of pigs and after

enrichment ...... 155

5.3.2 Screening for isolates with peptidase activity ...... 157

xii

5.3.3 Degradation of peptides involved in the etiology of celiac disease ...... 157

5.3.4 Characterization of cleavage fragments and proteolytic specificity ...... 160

5.3.5 Comparison with commercial probiotic strains ...... 163

5.4 Discussion ...... 163

5.5 Supplementary material ...... 167

5.6 References ...... 173

6. Chapter six: Conclusions, future directions and implications ...... 179

6.1 Conclusions and future directions ...... 179

6.2 Implications ...... 181

6.3 References ...... 183

REFERENCES ...... 186

Chapter one ...... 186

Chapter two ...... 200

Chapter three ...... 206

Chapter four ...... 211

Chapter five ...... 217

Chapter six ...... 223

xiii

LIST OF FIGURES

Figure 1.1- Word cloud representing the origin of lactobacilli...... 4

Figure 1. 2 Core genome phylogenetic tree of Lactobacillus sensu lato (Lactobacillus spp. and

Pediocccus spp...... 11

Figure 1.3 Genomic and metabolic characteristics of lactobacilli reflect different lifestyles ...... 12

Figure 1.4 Association of lactobacilli with host epithelia...... 19

Figure 1.5- Maximum likelihood trees comparing the phylogenetic structure of the (a) host- adapted species L. reuteri and the (b) nomadic L. plantarum ...... 28

Figure 1.6 Model of the evolution of lifestyles in the genus Lactobacillus ...... 30

Figure 2.1 Neighbor-joining tree of Lactobacillus reuteri ...... 63

Figure 2.2- Graphic representation of the experimental design...... 66

Figure 2.3 Cell numbers of L. reuteri in fecal samples (mice, pigs, humans) or cloacal swabs

(chickens) determined by quantitative culture ...... 67

Figure 2.4 Stacked bar plots showing the relative abundance of L. reuteri ...... 70

Figure 2.5 Ancestral state analysis ...... 76

Figure 2.6 Core and strain-specific gene content of L. reuteri lineage-Vi strains ...... 77

Figure 3.1 Structural organization of the TCS70529-30 genetic locus ...... 101

Figure 3.2 Growth curves of the parent strain L. reuteri 100 -23 and tΔhk70529 and Δrr70530

...... 102

Figure 3.3 Biofilm formation in the cell counts in the mouse forestomach ...... 103

Figure 3.4 4 In vitro biofilm formation and quantification of L. reuteri 100-23 and mutant strains

...... 104

Figure 3.5 Macroscopic and microscopic characteristics of in vitro biofilms ...... 106

Figure 3.6 SEM micrographs of L. reuteri 100-23 wild-type and mutant strains on polystyrene surfaces ...... 107

Figure 3.7 Resistance of wildtype and mutant in vitro biofilms to sheer stress ...... 108

xiv

Figure 4.1 Growth characteristics of L. reuteri ATCC PTA 6475 and ΔpduCDE ...... 126

Figure 4.2 Growth characteristics of L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE

(triangles) in a reconditioned E. coli spent medium ...... 131

Figure 4.3 Growth and metabolites produced by L. reuteri ATCC PTA 6475 (circles) and PTA

6475 ΔpduCDE (triangles) growing in a reconditioned E. coli spent medium ...... 132

Figure 4.4 Growth characteristics of L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE

(triangles) bMRS and in a reconditioned B. breve spent medium ...... 133

Figure 4.5 Growth and metabolites produce by L. reuteri ATCC PTA 6475 (circles) and PTA 6475

ΔpduCDE (triangles) growing in a reconditioned B. breve spent medium ...... 134

Figure 4.6 Illustrative and schematic representation 1,2 PD cross-feeding and glucose metabolism in the presence of 1,2 PD ...... 138

Figure 5.1 Experimental design ...... 156

Figure 5.2 Microbiological and analytical characterization of the bacterial enrichment cultures

...... 158

Figure 5.3 Degradation of the gluten peptides by the selected gluten peptide degrader GPD strains ...... 159

Figure 5.4 Characterization of peptide degradation by LC-MS/MS ...... 161

LIST OF TABLES

Table 1.1- Genomic and metabolic characteristics of species representing the different lifestyles of lactobacilli ...... 17

Table 2.1- L. reuteri strains used in the host adaptation assay ...... 57

Table 2.2 L. reuteri genomes used for phylogeny reconstruction and comparative genomics ..65

Table 2.3 Genes-specific to Poultry-VI and Human-VI strains ...... 78

Table 3.1 Bacterial strains and plasmids used in this study ...... 95

Table 3.2 Primers used in this study to generate knockout mutants ...... 97

Table 4.1 Bacterial strains used in this study ...... 121

Table 4.2 List of media used for cross-feeding experiments ...... 125

Table 4.3 Effect of addition of 1,2 PD on the end products of heterolactic fermentation of glucose by L. reuteri ...... 129

Table 5.1 Cleavage sites in the immunotoxic peptides by bacterial strain or strain mixture .... 162

xvi

GLOSSARY OF TERMS

Adaptation: Process by which an organism becomes more fitted to an environment as the result

of natural selection.

Allochthonous: Originates from a place other than that in which it is found.

Autochthonous: A true resident, found where formed.

Dispersal: Movements of individuals from a source location to another location where

establishment and reproduction may occur.

Free-living: Associated with plant material and/or environment without relying on an

eukaryotic host.

Habitat: The natural environment in which an organism lives.

Host-adapted: Specialized towards living in association with eukaryotic hosts, with adaptive

traits that facilitates persistence

Lactobacillus sensu lato: (From Latin: “in the broad sense”). Includes the lactobacilli and related

pediococci.

Lifestyle: The way of life of a species which allows its population to persist in nature.

Natural history: An organism's ecological interactions in its natural habitat and how they

evolved.

Niche (Hutchinsonian niche): Environmental conditions and resources within which a species

can maintain a viable population

Nomadic: Dynamic lifestyle that involves both environmental and host niches, with no signs

of specialization.

Specialized: Restricted in the breadth of its ecological niches as a result of trade-offs during

adaptation.

Symbiosis (From Greek: sym “with” and biosis “living”) Long-term associations between

genetically distinct organisms

xvii

1. Chapter One: Evolution and lifestyles of species of the genus Lactobacillus

1.1 Introduction

Lactobacilli are fastidious gram-positive bacteria that populate nutrient-rich habitats associated with food, feed, plants, vertebrate and invertebrate animals, and humans. Owing to their use in food, in biotechnology and in therapeutic applications, lactobacilli have substantial economic importance. Consequently, research focused on their role in food fermentations and spoilage (Chaillou et al. 2005; Gänzle and Ripari 2016; Stefanovic, Fitzgerald and McAuliffe 2017) biotechnological applications (Sun et al. 2015) and their functionality as ‘probiotics’, which are defined as “live microorganisms which when administered in adequate amounts confer a health benefit on the host” (Marco, Pavan and Kleerebezem 2006; Lebeer, Vanderleyden and De

Keersmaecker 2008; Bron, van Baarlen and Kleerebezem 2011; Hill et al. 2014). These studies have provided important information regarding the metabolism and functionality of a wide array of Lactobacillus species in the food environments and gastrointestinal tract, and their role in human and animal health. From an ecological and evolutionary perspective, however, these studies provide little insight as they are conducted in experimental settings that are abstracted from any natural history. Food habitats are man-made and date back less than 14,000 years

(Steinkraus 2002; Hayden, Canuel and Shanse 2013) which is short when considering that the natural history of lactobacilli with plants and animals dates back more than a billion years (Tailliez

2001; Battistuzzi et al. 2004). Furthermore, most probiotic research has been conducted with

Lactobacillus strains ‘allochthonous’ to the respective hosts in which they were studied (Walter

2008). We therefore lack information regarding the evolution of lifestyles in lactobacilli as it occurred in their true ecosystems in nature.

The genus Lactobacillus comprises more than 200 species that are characterized by a phylogenetic and metabolic diversity that exceeds that of a typical bacterial family (Sun et al.

2015). Recent phylogenetic analyses based on robust core genome phylogeny have revealed that lactobacilli can be subdivided into at least 24 phylogenetic groups (Zheng et al. 2015a); species of the genus Pediococcus form an integral part of the genus Lactobacillus. Accordingly, lactobacilli have been referred to as the Lactobacillus sensu lato including pediococci, or the

Lactobacillus Genus Complex to additionally include the related genera Weissella,

Leuconostoc, Oenococcus and Fructobacillus (Sun et al. 2015; Zheng et al. 2015a). The availability of genome sequences of lactobacilli has created a robust framework for large scale phylogenomic and comparative genomic analyses that can elucidate their evolution (Sun et al.

2015; Zheng et al. 2015a). In addition, population genomic and genetic analyses have allowed a detailed reconstruction of the evolutionary patterns in specific Lactobacillus species (Oh et al.

2010; Frese et al. 2011; McFrederick et al. 2013; Martino et al. 2016). If informed by an understanding of the metabolic traits of Lactobacillus groups and lineages, these analyses provide an opportunity to explore the ecological and evolutionary contexts in which these bacteria exist in nature and how their lifestyles have evolved.

This review compiles the available genomic and metabolic metadata for the genus

Lactobacillus to infer its evolution and natural history. Specifically a phylogenomic approach is applied to infer the natural habitat and then related to metabolic, functional and fine-scale phylogenetic analyses of model species. Lastly, remaining questions and how research in this dissertation aimed to address these questions is discussed.

1.2 Habitats of lactobacilli

Restricted by fastidious growth requirements, lactobacilli occupy nutrient-rich habitats which can be categorized into fermented or spoiled foods and animal feed, the environment including plants surface, soil, and the body of invertebrate and vertebrate animals (Fig. 1.1).

1.2.1.1 Food and feed

Lactobacilli dominate the microbiota of the vast majority of fermented foods and also occur as food spoilage organisms (Hammes and Hertel 2006; Gänzle 2015). Fermentation of silage, vegetables and many cereals relies on the microbiota of the raw materials as source of inoculum.

Other fermentations, including most dairy fermentations, sourdough, and fermented meats are controlled by back-slopping or “house microbiota” that are associated with the production environment (Scheirlinck et al. 2009; Su et al. 2012; Chaillou et al. 2013; Ripari, Gänzle and

Berardi 2016). Organisms in these fermentations are exposed to continuous propagation over decades or even centuries, essentially becoming domesticated to the fermentation environments

(van de Guchte et al. 2006; Vogel et al. 2011; Ding et al. 2014). Adaptation to conditions in food fermentations was suggested for L. delbrueckei ssp. bulgaricus, which shows rapid and ongoing reduction of the genome size (van de Guchte et al. 2006). However, genomic analysis of intestinal and sourdough isolates of L. reuteri indicated differential selective pressure in the two environments but not phylogenetic differentiation (Zheng et al. 2015b). The majority of the type strains of the Lactobacillus species have bee isolated from food (Fig1.1 a); however, food fermentations are unlikely to represent the primary habitat for Lactobacillus spp. (Fig. 1.1 b).

1.2.1.2 Environmental sites and plants

Lactobacilli occur frequently in sewage as a result of fecal contamination and occasionally in soils as part of the rhizosphere of plants or as a result of wash-off from the phyllosphere

(Kvasnikov, Kovalenko and Nesterenko 1983; Hammes and Hertel 2006). Despite the occasional reports of lactobacilli being isolated from wheat, beet and strawberries (Jacobs, Bugbee and

Gabrielson 1985; de Melo Pereira et al. 2012; Minervini et al. 2015), lactobacilli are a rare and minor component of the plant endophytes (Hallmann et al. 1997). Lactobacilli are detected in small numbers on plant surfaces, where traces of sugars can support their growth (Mercier and

Lindow 2000). Their numbers only increase upon damage of plant tissue when simple and

complex carbohydrates become available substrates (Müller and Lier 1994). The ecological role of plant-associated lactobacilli in nature is poorly understood, but because their occurrence is only sporadic, they are not considered plant symbionts but rather epiphytic (Stirling and

Whittenbury 1963; Mundt and Hammer 1968; Fenton 1987).

Figure 1.1- Word cloud representing the origin of lactobacilli.

The words describe the source of isolation of the type strains of lactobacilli; the square root of the font size of the words correlates to its frequency. (a) The isolation source of the 204 type strains of lactobacilli as described by Pot et al., (2014) or the newly described species. The description was simplified as follows: All strains of human or animal origin are designated as human or animal, irrespective of the site of isolation. The origin of all isolates from cereal mashes used for production of alcoholic beverages are designated as “mash”. The origin of all isolates from flowers, vegetable, sourdough, and silage fermentations were designated as “flower”, “pickle”, “sourdough” and “silage”, respectively, irrespective of the plant species. The origin of all strains isolated from kimchi, sauerkraut, and fermented cabbage was designated as “sauerkraut”. The origin of isolates from various stages of beer, wine, and apple cider fermentation was designated as “beer”, “wine”, and “apple”, respectively. The words “poultry” and “beef” represent meat; the words “chicken” and “cow” represent animals. (b) The origin of the same 203 type strains with a further simplification of the description of the origin as follows: the words representing spontaneous plant fermentations (pickle, sauerkraut and silage” was replaced by “plant”. The origin of all other food-associated organisms was omitted. The word cloud was generated with the online tool available at https://wordsift.org/.

1.2.1.3 Vertebrate and invertebrate hosts

Lactobacilli are reliably isolated from a variety of insects including flies and bees, and from vertebrates, particularly birds, rodents, humans and farm animals. The host range is likely larger as scientific investigations have been largely restricted to domesticated animals and humans

(Endo, Futagawa-Endo and Dicks 2010; McFrederick et al. 2013; Martino et al. 2016). Food storage organs such as the forestomach and crop appear to be the preferred habitat of lactobacilli in animal hosts. These organs are found in both insects (flies, bees, bumblebees) and vertebrate animals (poultry, rodents). In humans, lactobacilli are found in the oral cavity, gastrointestinal tract, with highest proportions in the small intestine and the vagina (Walter 2008).

1.3 What are the lifestyles of lactobacilli in nature?

Although we have a comprehensive knowledge of the origin of Lactobacillus strains, the precise ecological niches and lifestyles of these bacteria are difficult to unravel. To date, most functional research concerns the metabolic and, more recently, genetic adaptations to conditions that prevail in food and feed fermentations (Fig. 1.1a). However, although food fermentations provide opportunities for clonal expansion of specific species or phylogenetic groups (Cai et al.

2007; Chaillou et al. 2013; Zheng et al. 2015b), the adaptation of lactobacilli to these men-made habitats is coincidental and recent, and diversification, if it occurs, remains below the species level

(Cai et al. 2007; Chaillou et al. 2013; Zheng et al. 2015b). From an evolutionary perspective, food, feed, and biotechnological fermentations cannot be considered as habitats that supported speciation and cannot be considered for the elucidation of Lactobacillus lifestyles (Fig. 1.1b).

Although some species have been traced to animals, environment, and raw materials (Scheirlinck et al. 2009; Su et al. 2012; Chaillou et al. 2013; Ripari, Gänzle and Berardi 2016), the real ecological niches and natural history of most Lactobacillus species present in food and feed remains unknown.

Predictions about the exact natural history of lactobacilli are difficult even for organisms that are reliably found in habitats that support speciation. Lactobacilli can be ‘allochthonous’, meaning, they originate from a different place, and have, in contrast to ‘autochthonous’ species, neither an ecological nor evolutionary relationship with the habitat in which they are found. This concept is especially relevant for the gastrointestinal tract of humans where lactobacilli originate from fermented food (Tannock 2004; Walter 2008), but also relates to other habitats including wastewater, plants, flowers, and nectar, where lactobacilli may be present as fecal contaminants from vertebrates or insects. Autochthonous organisms establish long-term and stable populations of typical sizes and exert specific ecological functions in the habitat (Tannock 2004). However, even if such conditions are met, conclusions regarding the natural history of a species have to be drawn with caution. Allochthonous species establish stable populations when being introduced regularly into a habitat, and they may exert ecological functions even if such habitats are irrelevant to their evolution, as is the case of fermented foods. In addition, habitats or hosts that only allow sporadic and transient colonization may still play an important role in the overall lifestyle of a species, for example by providing vectors for dispersal or a temporal refuge (Vellend 2010). It conceivable that species possess a dynamic lifestyle comprised by more than one stable niche in which a classic autochthony could evolve.

Given these complexities, a combination of complementary approaches is required to reliably elucidate the natural history of lactobacilli. In the following sections the lifestyles of Lactobacillus species are deduced by synthesizing phylogenomic data with information on the metabolism of the bacteria, and inform these inferences with findings from more focused population genetics and functional studies. Specifically, (i) lifestyles are assigned based on a phylogenetic context, considering factors such as occurrence and frequency of detection/isolation as well as the strains’ metabolic characteristics and their ability to withstand environmental stressors present in given habitats; (ii) the evolutionary transitions between lifestyles are investigated by using a phylogenetic approach that is conceptually similar to that described by Sachs and co-workers

(Sachs, Skophammer and Regus 2011); (iii) patterns of genome evolution described to be associated with the evolution of symbiotic lifestyles are analyzed (Lo, Huang and Kuo 2016); (iv) this overview is then complemented with findings from fine-scale population genetic and functional studies on representative species that can serve as paradigms for the specific lifestyles represented within the lactobacilli.

1.3.1 Evolutionary insight through phylogenomics

The diversification of anaerobic clostridia and aerobic or facultative anaerobic bacilli and lactic acid bacteria roughly matches the “great oxidation event” that occurred ~2.5 billion years ago

(Battistuzzi et al. 2004). Lactobacillales then diverged from staphylococci and bacilli approximately 1.8 billion years ago (Battistuzzi et al. 2004), substantially predating the emergence of land plants (~500 million years ago), insects (~400 million years ago), mammals (~200 million years ago) and birds (~80 million years ago) (Shetty, Griffin and Graves 1999; Hedges et al. 2004;

Luo 2007; Clarke, Warnock and Donoghue 2011; Pires and Dolan 2012; Misof et al. 2014).

However, diversification within the genus Lactobacillus sensu lato likely intensified with the emergence and later diversification of the eukaryotic species with which lactobacilli became associated (Tailliez 2001).

To gain insight into lifestyle evolution of lactobacilli, we updated the core phylogenomic tree of Lactobacillus sensu lato (Zheng et al. 2015a) by adding species for which genome sequences became recently available (Fig. 1.2 and Table 1.S1). Based on isolation source, frequency of isolation, metabolic capabilities, growth temperature, and the ability to withstand environmental stressors present in given habitats, we assign species into three main lifestyle categories: free- living (encompassing environmental and plant isolates), host-adapted (associated with invertebrate or vertebrate hosts), or as ‘nomadic’ using the concepts proposed by Martino and co-workers (Martino et al. 2016). Remarkably, lifestyle assignments show a high correlation with phylogenetic grouping (Fig. 1.2). This association strongly suggests that monophyletic clades

within the lactobacilli are the results of adaptive evolution in different habitats, which resulted in the emergence of distinct lifestyles, with a high degree of phylogenetic niche conservation.

Specifically, the L. perolens, L. sakei, L. vaccinostercus, L. collinoides, L. brevis, and L. buchneri groups are almost completely composed of species that are rarely found in animals, and are therefore likely free-living. The species in the L. reuteri group are consistently associated with vertebrate hosts (human oral and vaginal cavity, intestinal tract, primates, other mammals, birds), while the L. salivarius group contains a monophyletic cluster associated with vertebrate hosts

(humans, rodents, birds, horses, cattle, swine, primates, whales) (Table S1) and a second cluster comprising mainly free-living species. The large and diverse L. delbrueckii group comprises clusters of species adapted to insects and to vertebrates including mammals such as pigs and hamsters and different species of birds). Species in the L. plantarum group and a cluster with the

L. casei group are nomadic, being reliably found in a wide variety of niches.

The conservation in the niche assignments of the deep-branching monophyletic lineages within the lactobacilli suggests that lifestyles evolved for long periods of evolutionary time and were stably maintained. These clear associations further pinpoint how lactobacilli evolved specific lifestyles. Lifestyle transitions occurred in 6 separate events (See Fig. 1.2 and legend for details).

The host adapted L. delbrueckii, L. salivarius, and L. reuteri groups likely evolved from free-living ancestors to become associated with vertebrates (events 1-3), while the L. fructivorans, L. kunkeei and L. mellifer groups evolved to become associated with insects (events 4 and 5). In the

L. delbrueckii group, a cluster of species related to L. apis appeared to have switched hosts and evolved from vertebrate-adapted to bee-adapted (event 6). In addition, L. fermentum is the only species in the L. reuteri group which is rarely found in intestinal ecosystems but frequently isolated from plants and spontaneously fermented cereals (Mundt and Hammer 1968; Hammes and Hertel

2006; Gänzle and Ripari 2016). L. fermentum could be an example a species undergoing reversion of the lifestyle from host-adapted to free-living, a process that has been documented

for environmental species that cluster within phylogenetic clades dominated by symbionts (Sachs,

Skophammer and Regus 2011).

1.3.2 Patterns of genome evolution reflect an evolutionary transition to a symbiotic lifestyle

The genomes of lactobacilli range in size from 1.27 (L. iners) to 4.91 Mb (L. parakefiri) and the number of genes between species varies considerably (Sun et al. 2015, Table S1). Lactobacilli underwent a process of genome reduction over the course of their evolution, losing on average approximately 3000 genes from the common ancestor and 1,300–1,800 genes in individual groups or species (Makarova et al. 2006; Sun et al. 2015; Zheng et al. 2015a). Gene decay in lactobacilli has led to substantial loss of functions in carbohydrate metabolism, amino acid and cofactor biosynthesis, leading to the fastidious nutritional requirements of the species (Makarova et al. 2006). This process is especially pronounced in lactobacilli associated with animals (Sun et al. 2015) and been attributed to nutrient-rich environments within host habitats (Makarova et al.

2006). However, genome reduction is an evolutionary process that is universally observed in symbionts and directly associated with the degree of host specialization (Lo, Huang and Kuo

2016). The stable environment provided by the host renders functions that were essential in the free-living ancestor redundant, which leads to an accumulation of loss-of-function mutations and pseudogenes followed by removal of these genetic regions, e.g. through mobile genetic elements

(Lo, Huang and Kuo 2016). Genome reduction is strongly correlated with host adaptation in

Lactobacillus species, genome size is significantly lower in host-adapted but not nomadic strains

(Fig. 1.3 a and b). Interestingly, genomes of host-adapted lactobacilli also show a reduction in GC content; this reduction of GC content is not observed in nomadic lactobacilli (Fig. 1.3 a and d).

This constitutes another well documented pattern observed in the genome evolution and is caused by strong mutational bias toward AT and non-adaptive loss of DNA repair genes of host-

adapted symbionts (Lo, Huang and Kuo 2016). Taken together, host-association in lactobacilli correlates with genomic events that are characteristic of the evolution of a symbiotic lifestyle.

1.3.3 Metabolic capabilities reflect lifestyle adaptations

Species within the Lactobacillus sensu lato show a substantial degree of variation in their metabolism (Gänzle 2015). The two phylogenetic clades of lactobacilli representing homofermentative and heterofermentative organisms, however, do not reflect association to specific habitats; both homo- and heterofermentative species associate with vertebrate animals, insects, or environmental habitats (Fig. 1.2). Remarkably, many habitats harbour both homofermentative and heterofermentative lactobacilli. Examples not only include intestinal habitats including the gut microbiota of fruit flies (L. plantarum and L. fructivorans groups), bees

(L. mellifer or L. delbrueckii and L. kunkeii groups, Anderson et al. 2013; Filannino et al. 2016) and the intestinal microbiota of vertebrate animals (L. delbrueckii and L. reuteri groups, Walter

2008) but also fermentation or spoilage microbiota in many foods including cereal fermentations, vegetable fermentations, and meat (Gänzle 2015; Hammes and Hertel 2006). Emerging evidence indicates that homo- and heterofermentative lifestyles are complementary rather than competitive

(Gänzle, Vermeulen and Vogel 2007; Tannock et al. 2012; Andreevskaya et al. 2016;

Andreevskaya 2017). Other differences in carbohydrate utilization patterns and growth temperature, however, provide helpful insights into niche adaptations. Free-living species are capable of growing at lower temperatures, while host-adapted species grow optimally at temperatures close to the body temperature of their corresponding hosts (Fig. 1.3e). The enzymatic repertoire of the species is also indicative of the substrates available in their natural habitats. Together, this information is essential to elucidate the exact lifestyle of the species and the characteristics of the niches to which the strains have adapted to.

The tree by was constructed according to Zheng et al. (2015) with the inclusion of 18 additional species for which genome sequence data became available since 2015. Eggerthia catenaformis was used as an outlier for the phylogenetic analysis. The inner segments delineate homofermentative and heterofermentative species, respectively. Members of the 24 phylogenetic groups are indicated by the same color for branches and the type strain of each group is printed in bold. Clusters in the L. delbrueckii and L. salivarius groups that differ in their ecology are separated by dashed lines. The solid circles in red represent genome sizes of these type strains; the area of the circle correlates with the

11 genome size. Color coding of the outer

ring indicates the lifestyle, if sufficient information is available. The habitat was assigned based on phylogenetic and ecological studies as well as literature related to the source of isolation; the assignment was additionally guided by database searches on the Integrated Microbial NGS Platform https://www.imngs.org (Lagkouvardos et al. 2016). Numbers indicate evolutionary transitions of lifestyle assuming an ancestral free- living state. Ancestral state reconstructions were executed in the Mesquite software package Version 3.2, http://mesquiteproject.org (Maddison and Maddisson 2017). Figure 1. 2 Core genome phylogenetic tree of Lactobacillus sensu lato (Lactobacillus spp. and Pediocccus spp.

Figure 1.3 Genomic and metabolic characteristics of lactobacilli reflect different lifestyles (a) Association between genome size and the number of coding sequences (CDSs). Pearson r = 0.95, p<0.0001. (b) Comparison of genome size (Mb) by lifestyle. (c) Association between GC content (%) and the number of CDSs. Pearson r = 0.58, p<0.0001. (d) Comparison of GC content (%) by lifestyle (E) Comparison of optimal growth temperature (mean ± SD) by lifestyle. Information was obtain from the genomes of type strains (Table S1). Representative species in panels A and C are color coded by lifestyle using the colors from Fig. 1.2. Box plots in panels B and D represent the median and the lower and upper quartiles. Whiskers extend to the last data point still within 1.5 inter-quartile range of the quartile. Kruskal–Wallis with a Dunn’s post hoc test was used to compare data between groups. Statistical significant groups are indicated (***, p<0.001; ****, p≤ 0.0001). Statistical analyses were performed in GraphPad Prism version 6.07 (GraphPad Software, La Joya, CA, USA).

1.4 Paradigms of Lactobacillus lifestyles

1.4.1 Free-living lifestyle

Species that are found in plant and environmental sources are scattered around the phylogenetic tree (Fig.1.2), which suggests an environmental ancestral condition for the most recent common ancestor of the genus. Free-living lactobacilli are clustered in the L. buchneri, and

L. collinoides groups, and all the species in the L. brevis, L. composti and L. perolens groups (Fig.

1.2).

Although it is difficult to determine if a lifestyle is strictly free-living, this lifestyle is strongly suggested by several characteristics of organisms in these clades. First, species within the phylogenetic groups are mostly isolated from plants or fermented plant products and very rarely from animals (Mundt and Hammer 1968; Daeschel, Andersson and Fleming 1987). Second, the metabolic and physiological properties of the strains are reflective of a free-living lifestyle. Most species within these groups are aerotolerant by using a Mn (II) defense mechanism against oxygen toxicity (Daeschel, Andersson and Fleming 1987). Additionally, their optimal growth temperature is closer to temperatures of terrestrial and aquatic habitats as most species are able to grow at 15°C - some even grow at 2-4°C – but not at 45°C (Table 1.S1 Fig.1.3f). Third, they possess large genomes (Fig. 1.3a and b) encoding a versatile range of enzymes to utilize a wide spectrum of substrates, including pentoses, sucrose, lactose, mannitol, melezitose, cellobiose, nitrate, citric acid, and malic acid (Danner et al. 2003; Zheng et al. 2015a; Martino et al. 2016).

Pentoses that are liberated upon degradation of plant materials as a result of hydrolysis of hemicellulose (Dewar, McDonald and Whittenbury 1963) are utilized by free-living lactobacilli through the pentose phosphate or phospoketolase pathways (Gänzle 2015). Interestingly, the ability to ferment pentoses is rarely found in yeast, suggesting a possible mechanism of niche partition between lactobacilli and yeast in their shared natural habitats (Mundt and Hammer 1968), which could be key to the success of lactobacilli in nature. Species that fit all three criteria well are L. hokkaidonensis and buchneri These species are isolated from grass silage, are aero-

tolerant, have a preference for pentose over hexose metabolism, an optimal growth temperature of 25 °C, and are psychrotrophic with a genome size of >2.3 kb (Tanizawa et al. 2015, Table 1.1)

1.4.2 Host-adapted lifestyle

The ability to colonize eukaryotic hosts benefits lactobacilli for several reasons; (i) their fastidious requirements for nutrients are satisfied in several host-associated niches; (ii) they often share the same food sources as the hosts (plants rich in simple and or complex carbohydrates); and (iii) they can use host animals as vectors to migrate to new habitats (Hammes and Hertel

2006; Mundt and Hammer 1968). Lactobacilli are found in vertebrates and insects. However, as described above, not all species isolated from animals are autochthonous, even those that differ markedly in the degree of specificity towards particular hosts or body sites, and the mechanisms by which symbiotic interrelationships are established and maintained. Examples are listed in

Table 1, and below research on representative species that can serve as paradigms for host- associated lifestyles in lactobacilli are discussed.

1.4.2.1 Lactobacilli adapted to vertebrate hosts

Species that colonize vertebrate hosts cluster within the L. delbrueckii, L. salivarius, and L. reuteri groups, are monophyletic and predominantly comprise host-associated species. This suggests that the vertebrate-associated lifestyle is the outcome of a long-term evolutionary process that brought about a stable co-existence with vertebrate animals. However, lineages did not remain within specific host species, and the members of the phylogenetic groups differ in terms of host range, colonization site (gut, oral cavity, vagina), and the degree of specialization.

This indicates that following initial adaptation to vertebrate hosts, further diversification and specialization occurred at the species level. Among the species for which the vertebrate lifestyle is best understood are L. reuteri, L. ruminis, L. salivarius, L johnsonii, L. amylovorus, and L. iners

(Table 1). A number of characteristics reflect the adaptation of these species to gastrointestinal

environments. They tolerate bile acids, are highly acid-resistant, and ferment oligo- and polysaccharides present in the diet of their host species (Kakimoto et al. 1989; Grill et al. 2000;

Lähteinen et al. 2010; Gänzle and Follador 2012; Ruiz, Margolles and Sánchez 2013; O’ Donnell et al. 2015; Zheng et al. 2015a; Krumbeck et al. 2016). Additionally, these species grow optimally at 37°C and higher (Table 1.1), which reflects the body temperatures of most mammals and birds.

Vertebrate-associated lactobacilli typically colonize a range of host species. Exceptions include the human vaginal species L. jensenii and L. iners, and the pig-associated L. amylovorus.

L. amylovorus is a dominant member of the porcine microbiota (Leser et al. 2002; Konstantinov et al. 2004, 2006; Chang et al. 2011; Kant et al. 2011) but is rarely detected in other animals

(Nakamura 1981; Guan et al. 2003; Reti et al. 2013) suggesting that it is host-specific to pigs. The species dominates the microbiota on the pars non-glandularis region of the pig stomach, which is characterized by a dense biofilm composed of lactobacilli (Pedersen and Tannock 1989; Mann et al. 2014). In addition, L. amylovorus is one of few lactobacilli capable of utilizing amylose by the extracellular hydrolysis of starch (Gänzle and Follador 2012), a trait that is likely to contribute to the ecological fitness of the species in the distal intestinal tract of pigs (Regmi et al. 2011).

The highest degree of niche specialization in vertebrate-adapted lactobacilli occurs in the human vagina. The vaginal microbiota is dominated by L. iners, L. crispatus, L. jensenii and L. gasseri (Anderson et al. 2014; Mendes-Soares et al. 2014). L. jensenii and L. iners are only found in this niche and the latter species shows the highest degree of specialization observed among the currently known lactobacilli. Compared to other all other lactobacilli, L. iners has a smaller genome and more complex nutritional requirements, reflected by its inability to grow on standard growth media (Macklaim et al. 2011; Petrova et al. 2016). L. iners has apparently evolved to an almost obligate symbiotic lifestyle that is highly dependent on the human host. The presence of specific genes, such as the Fe-S protein cluster involved in defense against oxidative stress from

H2O2 produced by other vaginal lactobacilli (Macklaim et al. 2011) also reflects specialization to the vaginal niche. Although biofilms are normally not observed in the healthy vagina, host

specificity of L. iners is likely achieved by specific adherence to epithelial cells in the vagina (Fig.

1. 4a; Macklaim et al. 2011).

The species L. reuteri, L. ruminis, L. johnsonii, L. salivarius, L. cripatus, L. acidophilus, and L. vaginalis have a broader host range and are found in different body sites (Table 1). However, the population structure of L. reuteri, L. ruminis, and L. johnsonii indicates that subpopulations within these species adapted and specialized to particular host animals. All three species separate in phylogenetic clusters that are highly reflective of host origin (Oh et al. 2010; Buhnik-Rosenblau et al. 2012; O’ Donnell et al. 2015). For L. reuteri, these clusters have been established by Amplified

Fragment Length Polymorphism, Multilocus Sequence Analysis (Oh et al. 2010 Fig. 1.5a), and whole genome phylogenies (Wegmann et al. 2015; Duar et al. 2017). The genome content of strains from different phylogenetic clusters is reflective of the niche characteristics in respective hosts (Frese et al. 2011). L. reuteri is regarded as autochthonous to the human gut (Reuter 2001) and has been found to be a prevalent member of the microbiota of traditional agriculturalist societies (Martínez et al. 2015). The genomes of human strains of L. reuteri are characterized by a closed pangenome and extensive deletion of large, adhesin-like surface proteins, but the ability to utilize glycerol and propanediol as electron acceptors, suggesting growth in the intestinal lumen

(Frese et al. 2011; Walter, Britton and Roos 2011). In contrast, rodent L. reuteri strains possess several large-adhesin like surface proteins and colonize by adhering to the surface of the squamous stratified epithelia of the forestomach of mice on which they form biofilms (Walter et al. 2005, 2007; Frese et al. 2013, Fig. 1.5a). Host specificity in L. reuteri has been experimentally demonstrated in competition experiments in gnotobiotic mice and more recently in chickens (Oh et al. 2010; Frese et al. 2011; Duar et al. 2017). L. reuteri isolated from both rats and mice cluster together and rat isolates are very competitive in mice. Similarly, isolates from chicken and turkeys group in the same phylogenetic lineages (Oh et al. 2010; Frese et al. 2011; Duar et al. 2017)

Table 1.1- Genomic and metabolic characteristics of species representing the different lifestyles of lactobacilli Mechanisms OTa Genome Group Organism Habitat GC (%) Lifestyle-associated traits of host References (°C) size (Mb) specificity Free-living

L. pentose fermentation, Tohno et al .(2013), vac Grass/silage 25 2.3 38.1 N/A hokkaidonensis aerotolerance Tanizawa et al. (2015)

Heinl et al. (2012) pentose fermentation, plant cell buc L. buchneri Grass/silage 37 2.5 44.4 N/A Kleinschmit et al. wall degradation (2006) Nomadic

Fruit flies; vertebrate bile resistance; metabolic Martino et al. (2016); pla L. plantarum digestive tract; plants and 37 3.2 44.5 versatility; two component N/A Siezen et al (2010) dairy products systems.; extracellular proteins

17 metabolic flexibility; adhesion to raw and fermented dairy; intestinal villi; bile resistance; silage, fermented Cai et al (2007, 2009); cas L. casei 30 2.8 46.5 environmental sensing and N/A vegetables, vertebrate Broadbent et al. (2012) adjustment; prototrophic to most digestive tract amino acids metabolic flexibility, fermentation raw and fermented dairy, Douillard of a wide range of carbohydrates; cas L. rhamnosus oral cavity, digestive tract 37 2.9 46.7 N/A (2013,2013a); Ceapa bile resistance; pili-mediated mucus of vertebrates, vagina (2015,2016); adhesion; immunomodulation. Vertebrate-adapted Digestive tract; predominant in the bile and acid resistance; motility, O’Donnell et al. (2015); sav L. ruminus bovine rumen; reported in 37 2.1 43.5 substrate foraging; Unknown Forde et al. (2011) humans, dogs, pigs, cats immunomodulation horses and primates. Proximal digestive tract; prevalent in rodents, pigs and chickens; reported in bile and acid resistance; adhesion Epithelial Oh et al (2011); Frese reu L. reuteri 37 1.9 38.6 humans, dogs, minks, and biofilm formation adherence et al. (2013) lambs, giraffes, cats and horses

bile and acid resistance; Digestive tract; prevalent extracellular amylases, surface- Kant et al. (2011); Grill del L. amylovorus in swine; reported in 37 2.0 37.8 Unknown attached "S-layers"; et al. (2001) chickens and horses. immunomodulation Human oral cavity and digestive tract.; reported in breast milk and vagina bile resistance, bacteriocin Raftis et at. (2011, sav L. salivarius 37 2.0 32.5 N/A and feces of pigs, production (Megaplasmid encoded) 2014); Li et a.l (2007) raccoons, chickens and hamsters Buhnik-Rosenblau et Proximal digestive tract of Bacteriocin production and bile del L. johnsonii 37 1.8 34.5 Unknown al. (2012); Pridmore rodents and poultry resistance (2004)

Fe-S - defense against peroxide. Epithelial Petrova et al. (2016); del L. iners Human vagina 37 1.3 32.5 Glycogen fermentation, adhesion adherence Macklaim et al (2011)

Insect-adapted

18 Adherence/ Ellegaard et al. (2015);

del L. apis Bee 37 1.7 36.6 biofilm formation in the hindgut Biofilm Anderson et al. (2013)

putative exopolysaccharide Ellagaard et al. (2015); formation, niche partition with mel L. mellis Bee 30 1.8 36.2 unknown Corby-Harris et al. other members of bee core (2014) microbiota Vojvodic et al.(2013), fructophilic, resistant to phenolics Anderson et al. (2013), kun L. kunkeei Flowers, grapes, bees 30 1.5 36.4 N/A and honey-desiccation Endo et al. (2013) Maeno et al (2016)

Figure 1.4 Association of lactobacilli with host epithelia. (a) Transmission electron micrograph image of immunogold- labeled L. iners cells in association with human vaginal epithelial cells, with L. iners cells indicated with an arrow (image from Macklaim et al. 2011). (b) Three dimensional confocal micrograph taken 24 hours after colonizing a germ-free mouse with a pure culture of the rat isolate L. reuteri 100-23. The specimen were stained with propidium iodide and imaged by confocal microscopy, which results in the bacterial cells to be colored red and the forestomach epithelium to appear green, as described by Frese et al. (2013). The image was taken by Christian Elowsky and Steven Frese at the University of Nebraska at Lincoln Microscopy Core. (c) Biofilm (red) composed of Lactic Acid Bacteria attached to a honeybee’s crop (green)(Vásquez et al. 2012). Images used under the Creative Commons Attribution (CC BY) license.

These findings demonstrate that L. reuteri has adapted to groups of related host species such as rodents or poultry that possess similar niches in their intestinal tracts and whose social behavior allows horizontal transfer of bacteria (Oh et al. 2010).

L. reuteri has been established as a model species to study mechanisms of host adaptation in lactobacilli (Walter, Britton and Roos 2011; Kwong and Moran 2015). Functional studies with loss-of-function mutations have demonstrated that the ecological success of rodent strains in the forestomach depends on biofilm formation (Fig. 1.4b), which is only observed in rodent strains, and resistace to gastric acidity (Walter et al. 2007; Frese et al. 2013; Krumbeck et al. 2016).

Inactivation of one single serine-rich surface adhesin specific to rodent strains with a devoted transport system (the SecA2-SecY2 pathway) completely abrogated biofilm formation, indicating that initial adhesion represented the most significant mechanism underlying host-specific colonization (Frese et al. 2013). Similar mechanistic studies are lacking in other species of lactobacilli but comparable genomic patterns of host adaptation are observed, e.g. for L. ruminis.

Human isolates of L. ruminis are aflagellate and non-motile while bovine, equine and porcine isolates are motile, with the latter two being hyper-flagellated (O’ Donnell et al. 2015). These differences in the expression of flagella and motility could reflect adaptation to the conditions in different hosts. Overall, the data available for L. reuteri and L. ruminis indicate that some lactobacilli evolved to a high degree of host-specialization. Moreover, robust clustering in defined phylogenetic groups based on host origin indicates that these host associations are maintained over evolutionary timescales. Finally, the high fidelity in epithelial recognition for biofilm formation of bacterial strains, as demonstrated for L. reuteri (Frese et al. 2013), provides a mechanism by which lineages are reliably transmitted from generation to generation and maintained over both ecological and evolutionary time scales.

Other host-adapted species appear to have a less specific and more ‘promiscuous’ lifestyle, both in terms of host range and body site. L. salivarius is indigenous to the human oral cavity

(Rogosa et al. 1953) and is one of few Lactobacillus species that has been consistently recovered

from the feces of human individuals for at least 18 months (Tannock et al. 2000). L. salivarius has also been obtained from breast milk (Martín et al. 2006) and a variety of body sites including the intestinal mucosa (Molin et al. 1993), tongue, rectum (Ahrné et al. 1998) and the vagina (Vera

Pingitore et al. 2009). This species is also found in pigs (Mackenzie et al. 2014), chicken

(Hammons et al. 2010) hamsters (Rogosa et al. 1953), and horses (Yuki et al. 2000). Phylogenetic analysis of strains from a variety of sources did not show clustering by origin. However, many isolates show signs of ongoing adaptation by genome decay (Raftis et al. 2011). L. vaginalis and

L. gasseri can be detected in oral and fecal microbiota of the same species (Dal Bello and Hertel

2006) and they are also members of the vaginal microbiota. Therefore, it appears that these species maintain more dynamic and flexible lifestyles regarding host range and ecological niche in comparison to L. reuteri and L. ruminis.

1.4.2.2 Lactobacilli associated with invertebrate hosts.

The association of lactobacilli with invertebrates as abundant members of the microbiome is a recent discovery (Shrivastava 1982; Engel and Moran 2013). Insect-associated species are distributed across the Lactobacillus phylogeny (McFrederick et al. 2012) (Fig 1.2) and cluster in phylogenetic groups with different levels of host specificity. Species associated with bees cluster in the L. kunkeei and L. mellifer groups and in the L. helsinborgensis clade of the L. delbrueckii group (Fig 1.2), which were termed as the Firm 4 and Firm 5 phylotypes prior to description of the species (Ellegaard et al. 2015). This finding suggests that association with the bee gut occurred in independent events (events 6 and 4, Fig. 1.2). Species of the L. fructivorans group (Fig. 1.2) are also often associated with bees but appear to be between species by floral transmission

(McFrederick et al. 2012).

Species belonging to all four groups are characterized by having small genome sizes (Zheng et al. 2015a; Maeno et al. 2016, Fig.1.2) and extremely limited carbohydrate fermentation capabilities (Ellegaard et al. 2015), being essentially restricted to a “sucrose and maltose diet”.

Heterofermentative lactobacilli associated with bees are fructophilic; they lack alcohol dehydrogenase activity and depend on the availability of fructose as electron acceptor (Endo,

Futagawa-Endo and Dicks 2009; Filannino et al. 2016; Maeno et al. 2016). Bee-associated lactobacilli in the homofermentative L. mellifer group share the restricted carbohydrate fermentation ability (Zheng et al. 2015a). These patterns of carbohydrate restriction are vastly different from vertebrate-adapted lactobacilli, which retain the ability to degrade a wider variety of carbohydrates despite their small genome size. It is likely that these differences reflect adaptations not only to the host’s diet (i.e. honey, nectar and pollen for bees) but also the differences in the competitive interactions that occur within the gut environments. Compared to vertebrates, bees harbor relative simple microbial communities composed of 9 bacterial species clusters, and there is compelling evidence that species occupy distinct and complementary metabolic niches within the bee gut (Powell et al. 2016).Therefore, specialization as a means of niche partitioning and syntrophic interaction seem to be one of the key mechanisms to the ecological success of bee-associated lactobacilli species (Kwong and Moran 2016).

Lactobacillus species are often dominant members of the microbiota of some species of

Hymenoptera (ants, bees, and wasps) (Kwong and Moran 2016). However, only honey and bumble bees have been described to date to harbor selective lineages of lactobacilli, suggesting a high degree of host-specificity in these hymenopteran hosts (McFrederick et al. 2013). Both the

L. mellifer group and L. helsinborgensis clade are almost ubiquitously represented in individual bees and are particularly abundant in adult workers and the queen bee, and individual lineages can be specific to honey and bumble bees (Vásquez et al. 2012; Kwong and Moran 2016). These species are oxygen-sensitive and have not been found outside the bee gut, and are likely obligate symbionts colonizing the anoxic regions of the distal hindgut. Genomic signatures of these species are in agreement with those of adapted symbionts (Lo, Huang and Kuo 2016). All species have small genomes, (< 2.1 Mb) with low GC contents ranging from 34.6 to a 36.6%. Most strains can grow at 15 °C and optimally at temperatures significantly lower than those adapted to

vertebrates (Fig 1.3e, Table 1.S1). The presence of genes involved in the utilization of trehalose; a disaccharide that is used for energy storage in insects, also emphasizes their adaptation to the insect gut (Ellegaard et al. 2015). Consistent with the adaptation to a sugar-rich environment, species of the L. helsinborgensis clade harbor a large number of PTS systems, carbohydrate transporters and a variety of modification enzymes including glucosidases, hydrolases, isomerases, racemases, epimerases, aldolases; more than most lactobacilli. Moreover, L mellifer and L. mellis encode strain specific genes with putative function in exopolysaccharide biosynthesis presumably involved in biofilm formation (Ellegaard et al. 2015).

Contrary to homofermentative bee-associated lactobacilli, the L. kunkeei group are dominant members in the crop microbiota of bee but can also be detected in pollen, nectar and hive materials, as well as from fresh flowers and fruits (Endo et al. 2012; Neveling, Endo and Dicks

2012; Anderson et al. 2013). L. kunkeei migrates frequently between honey bees and stingless bees and shows no evidence of specificity to either host, suggesting that the species is more

‘promiscuous’ than the members of the L. mellifer group and L. helsinborgensis clade (Tamarit et al. 2015). However, the genomic features and metabolic and biochemical traits of L. kunkeei clearly reflect adaptations to the bee gut and the hive environment (Tamarit et al. 2015). The phylogenetic group has small genomes (< 1.5 Mb), reduced GC content and grows at 15 °C (Fig.1.

3, Table 1.1). L. kunkeei is obligately fructophilic, osmotolerant and resistant to high concentration of phenolic acids present in pollen. Phenolic acids are used as electron acceptors (Filannino et al. 2016). Moreover, it resists the antimicrobial activity of royal jelly (Vojvodic et al. 2013) and the desiccant conditions in honey (Endo et al. 2012; Vojvodic et al. 2013). L. kunkeei is a major component of the biofilm that is found in the bee crop as determined by 16S rRNA sequencing

(Vásquez et al. 2012) (Fig. 1.4c). The ability of L. kunkeei to tolerate aerobic conditions is consistent with its capacity to migrate between hosts and the environment, such as fruits and flowers. However, the exact role of the environmental niches in the lifestyle of L. kunkeei is unclear. The species might be able to stably colonize fruits and flowers and/or uses them for

transmission. Alternatively, L. kunkeei might just be allochtnonous to fruits and flowers that gets deposited at these sites during bee pollination and foraging (McFrederick et al. 2012; Tamarit et al. 2015).

Species in the L. fructivorans group are shared between plants and insects including sweat bees and Drosophila. From its six species two were isolated from insects (L. fructivorans, L. vespulae), two from flowers (L. ixorae, L. florum), and two from fermented food products and therefore have an unknown environmental niche (L. sanfranciscensis and L. homohiochi) (Kitara,

Kaneko and Goto 1957; Endo et al. 2010; Vogel et al. 2011; Wong, Ng and Douglas 2011;

McFrederick et al. 2013; Techo et al. 2016, Fig. 1.S1). Although their lifestyle has not yet been studied, their reduced genome size (<1.5 Mb, with L. sanfransciscensis possessing one of the smallest genomes of all lactobacilli) and GC content (Fig. 1.3a-d) show the classic hallmarks of symbionts (Lo, Huang and Kuo 2016). In addition, species are either anaerobic or microaerophilic, reflecting adaptation to oxygen limiting conditions, which are likely to be found in the guts of insects but not in the environment (Pot et al. 2014, Table S1). We have classified this group as

‘insect-associated’ but more research is needed to elucidate the exact lifestyle of its members.

Contrary to bees, Drosophila species do not harbor a defined core-microbiota and the composition varies widely between wild and lab species. In fact, the microbiota of fruit flies is composed mainly of Acetobacter and Lactobacillus species found in other habitats, including the environment, dairy and vertebrate animals (Chandler et al. 2011; Wong, Ng and Douglas 2011;

Erkosar et al. 2013; Wong, Chaston and Douglas 2013). In addition, diet plays a major role in shaping the microbiome of Drosophila spp. (Wong et al. 2015). It appears therefore, that the association of lactobacilli with fruit flies is less host-restricted, more dynamic and includes the immediate environment of the insects (Wong et al. 2015). Such a lifestyle can be considered

‘nomadic’.

1.4.2.3 “Nomadic” species of lactobacilli

Most of the Lactobacillus species found in the human gut do not form stable populations and have been described as allochthonous as they are derived from food or feed (Tannock et al.

2000; Walter et al. 2001; Tannock 2004; Walter 2008). However, although not autochthonous in the classical sense, some Lactobacillus species, such as L. plantarum, L. casei, L.paracasei, and

L. rhamnosus possess adaptations to gut ecosystems and the oral cavity that allow them to persist for at least a limited time (Table 1.1). These species possess large genomes with little evidence for specialization to particular habitats, and they are found in vertebrate and invertebrate hosts and different body parts (gut, oral cavity, vagina), and in food materials, such as meat, fish, vegetables and raw or fermented dairy products (Kandler 1986; Stiles and Holzapfel 1997; Heilig et al. 2002; Wall et al. 2007; Delgado, Suárez and Mayo 2010; Siezen et al. 2010; Ceapa et al.

2016; Rossi et al. 2016). Recent research has provided convincing evidence that they represent examples of a nomadic lifestyle (Martino et al. 2016)

Nomadic Lactobacillus species cluster in two phylogenetic lineages, the L. plantarum group and a cluster within the L. casei group (Fig. 1.2). Their large genomes correspond to increased metabolic flexibility. Comparable to free-living lactobacilli, L. plantarum and L. casei retained the capacity for conditional respiration (Brooijmans, de Vos and Hugenholtz 2009; Zotta et al. 2016).

In addition, L. plantarum WCFS1 encodes a large spectrum of sugar uptake and utilization cassettes, allowing the organism to grow on numerous carbon sources. The genome of L. plantarum WCFS1 also encodes over 200 putative extracellular proteins, most of which are displayed in the cell surface. The presence of such proteins in plant-associated bacteria is often related to the degradation and utilization of plant oligo and polysaccharides, and likely contributes to the flexibility of L. plantarum to interact with its environment in different habitats (Siezen and van Hylckama Vlieg 2011). L. casei ATCC393, a strain isolated from cheese, is capable of synthetizing most amino acids (except for valine, leucine and isoleucine), and is thus able to thrive in protein-limited environments. It utilizes a great variety of carbohydrates (Cai et al. 2007) and

contains 16 two-components systems; the highest number observed among lactobacilli, indicating that gene expression is adjusted to diverse environments (Cai et al. 2009). The diversity of these three species has been extensively studied by both phenotypic and genotypic approaches

(Bringel, Curk and Hubert 1996; Torriani et al. 2001; De Las Rivas et al. 2005; Molenaar et al.

2005; Cai et al. 2007, 2009; Diancourt et al. 2007; Siezen et al. 2010; Broadbent et al. 2012;

Smokvina et al. 2013a; Martino et al. 2016). These approaches demonstrated high genetic and phenotypic intra-species diversity. Comparative genomic analysis of 54 L. plantarum strains demonstrated the absence of environmental specialization (Martino et al. 2016, and Fig. 1.5b), which had been already hypothesized in previous studies (Molenaar et al. 2005; Siezen et al.

2010). Genes involved in exopolysaccharide biosynthesis, sugar metabolism and the secretome showed the most variability amongst strains but did not relate to specialization to any specific habitats (Martino et al. 2016). Similarly, L. paracasei and L. casei did not show a correlation between the habitat and phylogenetic position as determined by core and pan-genome phylogenies coupled with analyses of variable regions (Cai et al. 2009; Smokvina et al. 2013a).

L. rhamnosus has been isolated from a large variety of habitats, including the human gastrointestinal tract, vaginal cavity, oral cavity and cheese, exemplifying its remarkable ecological adaptability. Comparable to L. plantarum and L. casei, L. rhamnosus likely resides in multiple niches, illustrating its nomadic lifestyle (Douillard et al. 2013a).

L. plantarum, L. casei, and L. rhamnosus do not form stable population in animal hosts but possess adaptive features to niches associated with humans and animals that contribute to their persistence (Walter 2008). For example, L. plantarum shows high tolerance to gastric juice and bile acids (Bron et al. 2004b; van den Nieuwboer et al. 2016). L. casei adheres to intestinal villi

(Galdeano and Perdigón 2004) and both L. casei and L. paracasei resist bile (Alcántara and

Zúñiga 2012) (Wang et al. 2010). L. rhamnosus possesses mucus-binding pili that might interact with the host epithelia in the oral cavity and the small intestine (Kankainen et al. 2009). L. plantarum responds to the gastrointestinal environment of mice by regulating a large array of

genes (Bron et al. 2004a). Interestingly, persistence of L. plantarum in the gastrointestinal tract of mice increases after only three passages (van Bokhorst-van de Veen et al. 2013). These studies suggest that some Lactobacillus species can adapt to intestinal ecosystems and temporarily persist despite not being autochthonous members of the resident microbiota.

Taken together, the findings indicate that some Lactobacillus species have evolved a nomadic lifestyle that exerts diverse selective pressures rather than promoting niche specialization.

Genomic and phenotypic characteristics of strains of these species appear unrelated to the origin of isolation, which highlights their ability to migrate across environments; in line with their ubiquitous presence and their ability to thrive on various substrates. This feature can be also seen as a strategy of dissemination, or from an ecological perspective, dispersal (Vellend 2010). During evolution, these species, originally associated with plants, may have developed the ability to inhabit the gut of animals feeding on plants, favoring dissemination to new habitats. Dispersal influences the dynamics, composition and structure of communities and the distribution and abundance of species. From an evolutionary perspective, it affects processes such as local adaptation, speciation and the evolution of traits that ultimately impact the natural history of species (Dieckmann, O’Hara and Weisser 1999). Nomadic Lactobacillus species could have evolved dispersal traits in the form of colonization factors of host animals, which allow these immotile bacterial species to disseminate. Nomadic lifestyles of lactobacilli have also been identified in insects such as some species of Hymenoptera (sweat bees and ants) and fruitflies,

(McFrederick et al. 2013; Matos and Leulier 2014) which represent excellent vectors for dissemination for bacteria that have their main habitat in plants and fruits. However, Lactobacillus lifecycle might even be more complex and dynamic, beginning with intestinal waste, followed by mechanical distribution to and among plants, and return to the host via the oral and alimentary cavity, as suggested in 1968 by Mundt and Hammer. Further studies should be directed to reconstructing the natural and evolutionary history of nomadic lactobacilli in both vertebrates and

invertebrates order to better understand their adaptation process and the relative importance of

free-living and host associated niches.

Figure 1.5- Maximum likelihood trees comparing the phylogenetic structure of the (a) host- adapted species L. reuteri and the (b) nomadic L. plantarum Tips of the branches are color coded by the strains' origin of isolation. The phylogeny of L. reuteri tree was inferred by multilocus sequencing analysis of 116 strains as described by Oh et al. (2010)

1.5 A model for the evolution of lifestyle transitions in the Lactobacillus

sensu lato

The synthesis of phylogenomic, metabolic, and data presented in this review provides a highly

consistent view on the evolution of distinct lifestyles of lactobacilli (Fig. 1.6). A free-living ancestry

for the Lactobacillus sensu lato is logical as symbioses with plants have not been described and the diversification from the bacilli predates the emergence of animals. From the ancestral state, the group has diversified and evolved lifestyles that cover the entire spectrum from free-living to strictly host-adapted, with a substantial variation in the reliance on environmental niches and the degree of host-specificity.

The phylogenomic data supports a model by which Lactobacillus lineages have diversified and evolved symbiotic lifestyles on five separate occasions (event 1-5 in Fig. 1.2), resulting in the

L. delbrueckii, L. salivarius, L. reuteri, L. mellifer, and L. kunkeei/L. fructivorans phylogenetic groups. This evolutionary process is reflected by adaptations to the host environment (bile and acid tolerance, growth at host body temperature, metabolic adaptations to insects) and genomic patterns (genome decay, decreased GC content, loss of biosynthetic enzymes) consistent with those found in other bacterial symbionts (Lo, Huang and Kuo 2016). Host-adapted lactobacilli differ in the degree of niche specialization and host dependence, and lifestyles can range from

‘promiscuous’ to completely host restricted, with L. iners representing the most extreme cultural representative. Selective epithelial adhesion (often followed by the formation of biofilms) appear to be a key mechanism by which lactobacilli maintain stable associations with hosts over evolutionary times as most animal sites with highly adapted species are characterized by adherent cells, e.g. the vagina, the crop of insects and birds, the forestomach of rodents and horses, and the pars esophagus in pigs (Fuller and Brooker 1974; Pedersen and Tannock 1989;

Tannock 1992; Yuki et al. 2000; Vásquez et al. 2012; Frese et al. 2013; Mann et al. 2014).

These host-adapted lifestyles likely evolved after ancestral plant, fruit, and flower associated lactobacilli became exposed to animals that were feeding on their primary habitats. Although this exposure was initiallly coincidental, bacterial traits that allowed the bacteria to tolerate the conditions in the host and allowed temporal persistence contributed to the transmission and hence, dispersal of lactobacilli. As such traits would ultimately increase the success of the lineages in their primary habitats, they could be shaped by natural selection even if they did not

allow stable colonization of the host, gradually increasing the relevance of host niches for the overall lifestyle. The result was the evolution of distinct and dynamic lifestyles that differ in the degree by which the microbes rely on environmental and host niches, and their dynamic interactions. Such ‘nomadic’ lifestyles remain represented within the genus Lactobacillus and might well constitute a transitional state from the free-living lifestyle to a specialized symbiosis.

Figure 1.6 Model of the evolution of lifestyles in the genus Lactobacillus Lifestyle evolution of lactobacilli from free-living to strictly host-adapted species. Representative species discussed in the text are included according to their lifestyle and their reliance on environmental niches and the degree of host-specificity.

1.6 Open questions

In symbiotic associations, both host and symbiont can reciprocally affect each other’s evolution (Moran 2006). When stably associated with a host, the bacterial symbiont is likely to adapt, which can lead to specialization and host-restriction. The ultimate result is an obligate symbiosis, in which the microbe depends on the host for survival. L. iners represents a classic example of such evolutionary process (Macklaim et al. 2011; Petrova et al. 2016). Rodent lineages of L. reuteri, although not yet obligate symbionts, have specialized to a degree that restricts their host range. However, so far it had remained unclear whether non-rodent lineages of L. reuteri have evolved to become host adapted, and open questions remain about the evolutionary relationships within lineage VI, which is shared by human and poultry isolates.

Chapter two describes a study that tested host adaptations of L. reuteri strains to swine, poultry and human volunteers.

Host anatomical sites where most host-adapted species colonize (bee cop, mouse forestomach and human vagina) have the characteristic of being lined by a stratified squamous tissue that allows lactobacilli to adhere and form biofilms (Fig. 1.4). These epithelia might therefore constitute an anatomical feature that evolved in the host to facilitate specific colonization of the beneficial symbionts. In mice, biofilm formation, in the forestomach is restricted to rodent strains of L. reuteri (Frese et al. 2013). In order to maintain such specificity, rigorous environmental sensing and selective epithelium recognition are likely to play a crucial role. Biofilm formation is often orchestrated in response to external signals recognized by signal transduction systems. Previously, a two component system (TCS) 70529-30 was identified to be specific to rodent strains, but its relevance in biofilm formation had remained unexplored (Frese et al. 2011).

Isogenic deletion mutants of the histidine kinase (hk70529) and the response regulator (rr70530) were generated in the rodent strain L. reuteri 100-23 and their ability to produce biofilms was

determined in monoassociated mice and further characterized using in vitro techniques. Findings are presented in Chapter three

Genes involved in adherence and biofilms formation are absent in most human strains of L. reuteri (Frese et al. 2011) thus suggesting a planktonic lifestyle for L reuteri in the human gastrointestinal tract. (Frese et al. 2011; Walter, Britton and Roos 2011). While lacking the genetic machinery to form biofilms, human strains conserve a 58-gene pdu-cbi-cob-hem cluster encoding functions in reuterin production, cobalamin biosynthesis, and glycerol and 1,2 propanediol (1,2

PD) utilization. Deoxy hexoses (fucose and rhamnose) derived from dietary and endogenous glycans are converted into 1,2 PD by a number of gut bacteria, making 1,2 PD readily available in the human gastrointestinal tract. It has been hypothesized that utilization of 1,2 PD might be a relevant colonization factor for L. reuteri in the human gut. Chapter four explores the ecological relevance of the pdu-cbi-cob-hem cluster in the context of cross-feeding of 1,2 PD produced from the fermentation of fucose and rhamnose by two different gut bacteria.

Findings from probiotic research suggest that host-specificity might not a be requisite for lactobacilli to provide beneficial effects upon the host. The list of Lactobacillus species with recognized probiotic properties comprises host-adapted species such as L. reuteri, L. johnsonii and L. acidophilus, as well as nomadic and free-living organisms such as L. plantarum, L. casei and L. fermentum (Floch et al. 2015). This indicates that lactobacilli, irrespective of lifestyle, can be introduced into the gastrointestinal tract and remain viable. Nonetheless, is it a logical working hypothesis that those adapted to gastrointestinal environments will remain physiologically and are more likely to influence host physiology. Work in Chapter 5 presents a unique approach to exploit the metabolic capacity of lactobacilli adapted to the vertebrate gut for therapeutic purposes.

Strains of the host-adapted species L. ruminis, L. johnsonii, L. amylovorus, and L. salivarius

isolated from the porcine gut, are characterized by their ability to degrade meta-stable peptides involved in the pathogenesis of celiac disease.

Finally, conclusions from research presented in this thesis and the implications for future studies are presented in Chapter 6.

1.7 Supplementary material

Table S1.1 Metadata of type strains with lifestyle assignment in Fig. 1.2

L.) silage L.)

Phleum pratense

cow dung

secondary incheese flora aged

apple mash

broiler leg broiler

timothy grass ( timothy grass

must of BobaL. grape variety of grape must BobaL.

sap of Quercus sp.

apple juice press cider from

malted barley malted

surface of a eutrophic freshwater pond ofsurface eutrophic a freshwater

Moto starter of sake Moto starter

fermented meat product meat fermented

grass silage grass

vacuum-packaged beef

milk

fermented dairy beverage fermented

orange lemonade orange

chinese traditionaL. fermented vegetable Suan cai vegetable fermented chinese traditionaL.

flower of palustris Caltha flower

silage

composting materiaL. of distilled shochucomposting residue materiaL.

fermented cane molasses at alcohoL. at plants cane molasses fermented

brewery environment brewery

fermenting apple juice fermenting

fermented grains grains fermented

kefir grains kefir

non-salted turnips) pickle of red and solution stems leaves used in production fermented of sunki traditionaL. (Japanese

composting materiaL. of distilled shochuncomposting materiaL. residue

human saliva

non-salted turnips) pickle of red and solution stems leaves used in production fermented of sunki traditionaL. (Japanese

wine

composting materiaL. of distilled shochucomposting residue materiaL.

maize silage maize

tomato pulptomato

artisanal wheat sourdough wheat artisanal

rice sourdoughrice

pickles

brewery

senmaizuke

sourdough, manufactured with wheat, rye and flour spelt rye sourdough, with wheat, manufactured

Kimchi

wheat sourdough wheat

human faeces

Source of Isolation

feces-vertebrate

dairy

fruits

meat

silage

fruits

plants/veretables plants/veretables

fuits

fermented cereal fermented

aquatic sample

fermented cereal fermented

meat

silage

meat

dairy

fuits

fermented vegetables fermented

plants/flowers

silage

fermented cereal fermented

plants/vegetables/food

fermented cereal fermented

fermented fruits fermented

fermented food fermented

fermented vegetables fermented

fermented cereal fermented

human saliva

fermented vegetables fermented

fermented fruits fermented

fermented cereal fermented

silage

fruits

fermented cereal fermented

fermented vegetables fermented

fermented cereal fermented

fermented vegetables fermented

plants/vegetables

food/enviromen

Main Habitat

NZ_AYYY00000000.1

NZ_AWTT00000000.1

NZ_BACO00000000.1

NZ_AZFE00000000.1

NZ_AP014680.1

NZ_AZEG00000000.1

NZ_BALC00000000.1

NZ_BACP00000000.1

NZ_AZDX00000000.1

NZ_AYZD00000000.1

NZ_BALW00000000.1

NZ_AZFG00000000.1

NZ_AYZB00000000.1

NZ_BAMJ00000000.1

NZ_AZDL00000000.1

NZ_AVAA00000000.1

NZ_AZEC00000000.1

NZ_AUEH00000000.1

NZ_AYZL00000000.1

NZ_AYYK00000000.1

NZ_BAMK00000000.1

NZ_BBAD00000000.1

NZ_BBAG00000000.1

NZ_BBEQ00000000.1

AZEN00000000.1

NZ_AYYV00000000.1

NZ_AZEA00000000.1

NZ_AZEI00000000.1

NZ_BBAR00000000.1

NZ_AZGK00000000.1

NZ_AZED00000000.1

NZ_AZEB00000000.1

NZ_ACGP00000000.1

NZ_BAKI00000000.1

NZ_AZEY00000000.1

NC_018610.1

NZ_AZDW00000000.1

NZ_AZFC00000000.1

NZ_AYZH00000000.1

NZ_JQCA00000000.1

NZ_ARTH00000000.1

NZ_AZDT00000000.1

NZ_CP012033.1

NZ_AZFS00000000.1

NZ_AWVK00000000.1

Genome Accession No.

2359

1713

2339

1726

2233

2407

2039

2486

2127

2117

1646

1886

1692

1791

1699

2862

3003

2894

1210

1667

2822

2574

2435

2313

4704

2007

2488

2550

2232

2311

2204

2647

2336

2204

2906

2288

2392

2410

2064

2129

2314

2162

2608

2506

2291

CDSs

DSM 20634

DSM 29958

DSM 5007

DSM 15707

DSM 26202

DSM 19971

DSM 21376

DSM 20444

DSM 19519

DSM 21051

DSM 20017

DSM 15831

DSM 20719

DSM 14340

DSM 20019

LY 73 LY

DSM 12744

DSM 16991

DSM 23037

DSM 20335

DSM 18527

DSM 23365

DSM 15502

DSM 20515

DSM 10551

DSM 20587

DSM 19904

DSM 19907

DSM 18390

DSM 5707

DSM 19908

DSM 19906

DSM 20176

DSM 18382

DSM 14421

DSM 20057

DSM 19395

DSM 15429

DSM 21775

DSM 22467

ATCC 53295ATCC

DSM 19117

JCM JCM 16448

DSM 16381

ATCC 14869ATCC

Type Strain

43.5

39.8

39.0

35.5

38.1

36.9

38.4

36.0

34.8

37.4

41.0

40.2

41.8

42.0

56.4

49.1

53.1

34.5

38.0

43.9

46.9

46.8

46.1

41.6

41.7

42.0

43.0

45.2

43.4

42.4

41.7

39.6

42.1

40.0

44.4

53.6

55.9

48.6

49.1

49.0

52.0

49.2

49.4

45.9

(%)

content

2.6

1.9

2.7

1.8

2.3

2.7

2.5

2.7

2.3

2.4

1.9

2.0

1.8

2.1

1.8

3.2

3.3

3.1

1.3

1.8

3.4

3.5

3.6

4.9

2.7

2.9

3.1

2.6

2.3

3.0

2.6

2.8

3.3

2.5

2.7

2.2

2.4

2.6

2.5

3.0

2.8

2.5

size (Mb)

Genome

microaerophilic-anaerobic

anerobic

microaerophilic

Microaerophilic

facultative anaerobic

Microaerophilic

facultative anaerobic

microaerophilic

facultative anaerobic

microaerophilic

facultative anaerobic

microaerophilic

facultative anaerobic

microaerophilic

facultative anaerobic

microaerophilic

facultative anaerobic

microaerophilic

microaerophilic-anaerobic

anerobic

microaerophilic

facultative anaerobic

microaerophilic

aerobic

Atmosphere

U to 42to U °C

n/a

at 45 °C

Growth

Y (growth at 4 at °C) (growth Y

Y (growth at 2-4 at (growth Y °C)

n/a

Y (growth at 10 at (growth Y °C)

°C

Growth at 15

(°C)

temperature

growth

Optimal

L. vaccinostercus L.

L. wasatchensisL.

L. suebicusL.

L. oligofermentansL.

L. hokkaidonensisL.

L. uvarumL.

L. sucicola L.

L. mali L.

L. hordeiL.

L. aquaticusL.

L. sakeiL. subsp. sakei

L. sakeiL. subsp. carnosus

L. graminisL.

L. fuchuensisL.

L. curvatus L.

L. shenzhenensisL.

L. perolensL.

L. harbinensisL.

L. floricola L.

L. dextrinicus L.

L. composti L.

L. similis L.

L. paracollinoidesL.

L. collinoidesL.

L. parakefiriL.

L. kefiri L.

L. sunkiiL.

L. rapi L.

L. parafarraginisL.

L. parabuchneriL.

L. otakiensisL.

L. kisonensisL.

L. hilgardiiL.

L. farraginisL.

L. diolivoransL.

L. buchneriL.

L. zymae L.

L. spicheri L.

L. senmaizukeiL.

L. paucivoransL.

L. parabrevisL.

L. namurensisL.

L. koreensis L.

L. hammesii L.

L. brevis L.

Free-living

Species

vac

sav

sak

per

dex

com

col

buc

bre

bre Group

Table S1.1 (cont.). Metadata of type strains with lifestyle assignment in Fig. 1.2

alfalfa alfalfa silage

peony (Paeonia suffruticosa)

Portuguese traditional dry fermented sausage fermented dry traditional Portuguese

San Francisco sourdough

commerciaL. grape wine undergoing a sluggish/stuck undergoing a wine alcoholic grape commerciaL. fermentation

bees

Inula ciliaris var. glandulosa,Inula chrysanthemum ciliaris a var.

the honey mellifera Apis honeybee stomachthe of the

honey stomach of honey bee (Apis mellifera mellifera) honey stomach of (Apis mellifera honey bee

bee

honey stomach of honey bee (Apis mellifera mellifera) honey stomach of (Apis mellifera honey bee

bees

the honey mellifera Apis honeybee stomachthe of the

bumble bumble bees

Corn silage Corn

Pickled cabbage

fermented cassava roots (fufu) cassava roots fermented

cocoa fermentation heap bean

beer contaminant beer

fermented radish fermented

Chinese pickleChinese

cheese

pasteurized milk pasteurized

Source of Isolation

silage

plant/flower

fermented meat fermented

fermented cereal fermented

bee, plant/flower/fermented fruit plant/flower/fermented bee,

bee

plant/flower

bee

silage

fermented vegetable fermented

fermented plant fermented

fermented cereal fermented

fermented vegetable fermented

dairy

Main Habitat

NZ_AEQY00000000.1

NZ_AYZI00000000.1

NZ_JQBN00000000.1

NC_015978.1

NZ_AZCK00000000.1

NZ_JXCT00000000.1

NZ_AYYQ00000000.1

NZ_JXJQ00000000.1

NZ_JXBZ00000000.1

NZ_JXJR00000000.1

NZ_JXLH00000000.1

NZ_JXLI00000000.1

NZ_JXLG00000000.1

NZ_JXBY00000000.1

FOMN00000000.1

NZ_CM001538.1

NZ_ACGZ00000000.2

AZFR00000000

NZ_AZCQ00000000.1

NZ_AYGX00000000.2

NZ_CP013130.1

NZ_LFEE00000000.1

NZ_AZGH00000000.1

NZ_AZCO00000000.1

NZ_AYYJ00000000.1

Genome Accession No.

1283

1273

1301

1234

1328

1292

1402

1654

1622

1770

1955

1959

1553

1883

1570

2915

2892

3013

2619

2973

2751

2627

2779

2566

2059

CDSs

DSM 20203

DSM 22689

DSM 20571

DSM 20451

DSM 12361

DSM 26257

DSM 23829

DSM 26254

DSM 26255

DSM 26265

DSM 26263

DSM 26256

LMG 26964LMG

DSM 26262

DSM 28793

DSM 20314

ATCC 14917ATCC

DSM 16365

DSM 20021

DSM 21115

DSM 10667

DSM 20350

DSM 5622

DSM 20011

DSM 20258

Type Strain

38.9

41.1

38.8

34.7

36.4

34.5

31.9

39.3

36.2

36.4

35.8

35.7

36.6

35.5

34.6

46.4

44.5

45.0

46.7

45.0

44.0

43.5

46.5

46.4

(%)

content

1.4

1.3

1.4

1.3

1.5

1.4

1.5

1.8

2.0

2.1

1.7

2.1

1.7

3.4

3.2

2.9

3.3

3.1

2.9

2.8

2.4

size (Mb)

Genome

microaerophilic

microaerophilic - anaerobic

microaerophilic

anaerobic

microaerophilic

anaerobic

microaerophilic

Atmosphere

some

at 45 °C

Growth

Y (growth at 10 at °C) (growth Y

°C

Growth at 15

(°C)

temperature

growth

Optimal

"Firm 5" "Firm

"Firm 4" "Firm

"Firm 5" "Firm

L. fructivorans L.

L. florum L.

L. homohiochiiL.

L. sanfranciscensis L.

L. kunkeeiL.

L. apinorumL.

L. ozensis L.

L. mellifer mellifer L.

L. mellis mellis L.

L. helsingborgensisL.

L. kimbladiiL.

L. melliventris melliventris L.

L. apisL.

L. kullabergensisL.

L. bombicolaL.

Insect-adapted

L. pentosusL.

L. plantarumL. subsp. plantarum

L. plantarumL. subsp. argentoratensis

L. rhamnosusL.

L. fabifermentansL.

L. paraplantarumL.

L. herbarum L.

L. paracasei subsp.L. paracasei

L. casei L.

L. paracasei subsp.L. tolerans

Nomadic

Species

fru

kun

mel

del

pla

cas

pla

cas

cas Group

Table S1.1 (cont.) Metadata of type strains with lifestyle assignment in Fig. 1.2

)

Ziphius cavirostris

faeces of faeces chicken

intestine of chickenintestine

beaked whales ( whales beaked

dentaL. plaque of baboondentaL.

saliva

municipaL. sewage

pig faeces

feces, wild Japanese wood wild mouse Japanese feces,

intestine of rat intestine

bovine rumen

faeces of (horse) faeces thoroughbred

fermented fish fermented

faeces of faeces horses

pig smalL. intestine pig smalL.

gorilla

feces offeces healthy a 100-year-old female Japanese

gastric biopsies, humangastric stomach mucosa

rye-bran sourdough rye-bran

pigeon, crop

gastric biopsies, humangastric stomach mucosa

intestine of adultintestine

human vagina

pigs and chicken

sourdough

human saliva

vagina. swab from patient with trichomoniasis patient vagina. swab from

chickens

human urine

lung of parrot

human intestine

chicken crop

gastric biopsies, humangastric stomach mucosa

human vagina. discharge

intestine of rat intestine

gastric biopsies, humangastric stomach mucosa

healthy thoroughbred racehorse healthy thoroughbred

eye

chicken crop

cattle / pigscattle

chicken, intestine

silage cattle feed cattle silage

human blood

chicken crop

acidified beer wort acidified beer

faeces of faeces hamster

thoroughbred horses thoroughbred

Source of Isolation

chicken

whale

baboon

human saliva

sewage

pig

mouse

rat

cattle

horse

fermented fish fermented

horse

pig

gorilla

human

fermented cereal fermented

pigeon

human

human, rodent, chicken,pighuman, rodent,

human

pig, chicken

fermented cereal fermented

human

chicken

human vagina

parrot

human

chicken

human

rat

human

horse

human vagina

chicken

cattle/pig

chicken

silage

rodent

chicken

pig, fermented cereal pig, fermented

hamster

horse

Main Habitat

NZ_AYZA00000000.1

NZ_AYYZ00000000.1

NZ_AUHP00000000.1

NZ_AEOF00000000.1

NZ_ACGT00000000.1

NZ_CVQY00000000.1

NZ_AUHQ00000000.1

NZ_AZFT00000000.1

NZ_AYYN00000000.1

NC_015975.1

NZ_BAML00000000.1

NZ_AZFI00000000.1

NZ_BAMI00000000.1

NZ_AZEQ00000000.1

BCAH01000001 BCAH01000063.to

NZ_AYZR00000000.1

NZ_ACLL00000000.1

NZ_AZER00000000.1

NZ_AZFK00000000

NZ_AZFN00000000.1

NZ_AZDD00000000.1

AZEW01000000

NZ_AZGO00000000.1

NZ_AZGM00000000.1

AZGE00000000

NZ_ACGV00000000.1

NZ_ACLM00000000.1

NZ_ACLN00000000.1

NZ_AUEI00000000.1

NZ_CAKE00000000.1

NZ_CAKD00000000.1

NZ_AZFM00000000.1

NZ_ACGQ00000000.2

NZ_AZGN00000000.1

NZ_ACGU00000000.1

NZ_AZDU00000000.1

NC_014106.1

NZ_CAKC00000000.1

NZ_AZCM00000000.1

NZ_AZFU00000000.1

NZ_AYZG00000000.1

NZ_ACGR00000000.1

NZ_AZEL00000000.1

NZ_BALQ00000000.1

NZ_ADNY00000000.1

NZ_BALY00000000.1

NZ_BBAS00000000.1

Genome Accession No.

1529

1370

1236

1685

1858

1960

1621

1913

1938

1836

1458

2002

1878

2067

1582

1528

2031

1611

1993

1727

1846

1570

1502

1753

1826

1634

1913

1165

1308

1773

1718

1859

1363

1730

1947

1825

1840

1745

1805

1749

1662

1575

1760

1567

1406

1581

1186

CDSs

DSM 20655

DSM 20653

DSM 22408

DSM 20602

DSM 20555

DSM 20509

DSM 16049

DSM 16634

DSM 20452

DSM 20403

DSM 18933

DSM 15836

DSM 15833

DSM 13345

DSM 28356

DSM 24302

DSM 16041

DSM 13145

DSM 15946

DSM 16045

DSM 20016

DSM 14060

DSM 8475

DSM 6035

DSM 4864

DSM 5837

DSM 20075

DSM 13335

DSM 15354

DSM 23910

DSM 23907

DSM 16043

DSM 20557

DSM 6629

DSM 16047

DSM 19284

DSM 20584

DSM 23908

DSM 20531

DSM 16761

DSM 21401

DSM 10533

DSM 10532

ATCC 33323ATCC

DSM 11664

DSM 5661

DSM 18793

Type Strain

40.1

33.7

41.1

32.5

41.7

42.5

38.6

40.0

43.5

34.0

39.1

39.0

46.4

48.1

39.1

51.1

42.5

50.9

41.6

38.6

40.8

54.4

48.1

50.0

40.6

36.8

32.5

35.7

35.1

38.6

36.1

34.4

35.3

36.0

47.7

36.9

37.8

37.5

34.0

34.5

36.4

35.0

38.2

35.0

42.6

(%)

content

1.7

1.5

1.4

1.9

2.0

2.1

1.7

2.1

2.2

2.1

1.7

2.3

2.2

2.3

1.6

2.2

1.7

2.0

1.8

1.9

1.7

2.0

1.8

2.2

1.3

1.5

1.9

2.1

1.6

2.0

2.2

2.1

2.0

1.9

2.0

1.9

1.8

1.9

1.8

1.5

1.8

1.6

size (Mb)

Genome

anaerobic

microaerophilic

microaerophilic - anaerobic

anaerobic

microaerophilic-anaerobic

n/a

anaerobic

microaerophilic

microaerophilic - anaerobic

microaerophilic

microaerophilic-anaerobic

microaerophilic

microaerophilic-anaerobic

anaerobic

microaerophilic

anaerobic

microaerophilic - anaerobic

anaerobic

microaerophilic

anaerobic

microaerophilic - anaerobic

microaerophilic

anaerobic

microaerophilic

anaerobic

microaerophilic

microaerophilic - anaerobic

microaerophilic

aerobic

Atmosphere

N (max 42°C) (max N

at 45 °C

Growth

Y Y (slow)

W (weak) W

°C

Growth at 15

(°C)

temperature

growth

Optimal

L. aviariusL. subsp. aviarius

L. aviariusL. subsp. araffinosus

L. ceti L.

L. animalisL.

L. salivariusL.

L. agilisL.

L. saerimneri L.

L. apodemiL.

L. murinusL.

L. ruminis L.

L. hayakitensisL.

L. acidipiscisL.

L. equiL.

L. mucosae L.

L. gorillaeL.

L. senioris L.

L. antri L.

L. frumenti L.

L. ingluvieiL.

L. gastricus L.

L. reuteri L.

L. coleohominisL.

L. pontisL.

L. panisL.

L. oris L.

L. vaginalisL.

L. helviticusL.

L. iners L.

L. psittaci L.

L. hominisL.

L. pasteurii L.

L. kalixensisL.

L. jenseniiL.

L. intestinalisL.

L. ultunensisL.

L. equicursoris L.

L. crispatus L.

L. gigeriorumL.

L. amylovorusL.

L. kitasatonisL.

L. taiwanensisL.

L. johnsoniiL.

L. gallinarumL.

L. gasseri L.

L. amylolyticusL.

L. hamsteri L.

L. equigenerosiL.

Vertebrate-adapted

Species

sav

reu reu

reu

del

del Group

10-1 10-1 L. helsingborgensis L. vini L. helveticus L. hordei 10-2 L. harbinensis L. salivarius 10-2 L. casei L. animalis L. sakei L. plantarum 10-3 10-3

10-4 10-4

10-5 10-5

10-6 10-6

10-7 10-7

10-1 10-1 L. rossiae L. siliginis L. vespulae 10-2 L. fermentum L. sanfranciscensis 10-2 L. reuteri L. senioris L. kunkeei L. buchneri 10-3 L. parakefiri 10-3

10-4 10-4

10-5 10-5

10-6 10-6

10-7 10-7

frequency of sequences with 99% identity / of identity totalsequences with of number sequences frequency 99%

bee bee

plant plant

food food

insect insect

rodent rodent

piggut piggut

fish gutfish gutfish

chicken chicken

primates primates

bovinegut bovinegut

humangut humangut

humanoral humanoral

humanvagina humanvagina Origin of metagenome dataset

marineinvertebrates marineinvertebrates

Figure S1. Frequency of selected Lactobacillus species in food, animals, and the environments. Full length 16S rRNA sequences of the species shown were used as query sequence to search the metagenome database at www.imngs.org. In January 2017, the database contained 88580 metagenome datasets with a total of 2,565,966,305 sequences; the selected categories shown represent 1,014,194,428 sequences. Colours represent lactobacilli adapted to vertebrate hosts (dark red), insects (light red), nomadic lactobacilli (green), or environmental lactobacilli (blue). Lactobacilli with unknown habitat are without colour; hatched bars represent food-fermenting lactobacilli. The category “food’ includes the IMNGS classifications “food” and “fermentations”; the category “rodent” includes “mouse gut and skin”, “rat gut” and “rodent”; the category “insect” includes “insect gut”, “mosquito” and “termites”; the category “marine invertebrates” includes “coral metagenome”, “echinoderm”, “jellyfish”, “mollusc”, “sea squirt”, “shrimp” and “sponge”; the category “plant” includes “compost”, “endophyte”, “leaf”, “phyllosphere”, “plant”, “root and associated fungi”, and “shoot”. Other categories were omitted because the frequency of sequences matching Lactobacillus sequences was below 10-5, or because the sequences were obtained from unclassified or artificial habitats.

1.8 References

Ahrné S, Nobaek S, Jeppsson B et al. The normal Lactobacillus flora of healthy human rectal and oral mucosa. J Appl Microbiol 1998;85:88–94.

Alcántara C, Zúñiga M. Proteomic and transcriptomic analysis of the response to bile stress of Lactobacillus casei BL23. Microbiology 2012;158:1206–18.

Anderson AC, Sanunu M, Schneider C et al. Rapid species-level identification of vaginal and oral lactobacilli using MALDI-TOF MS analysis and 16S rDNA sequencing. BMC Microbiol

2014;14:312.

Anderson KEK, Sheehan THTTH, Mott BMB et al. Microbial Ecology of the Hive and

Pollination Landscape: Bacterial Associates from Floral Nectar, the Alimentary Tract and Stored

Food of Honey Bees (Apis mellifera). Gerardo NM (ed.). PLoS One 2013;8:e83125.

Andreevskaya M. Ecological fitness and interspecies interactions of food-spoilage-associated lactic acid bacteria: insights from the genome analyses and transcriptome profiles. 2017.

Andreevskaya M, Johansson P, Jääskeläinen E et al. Lactobacillus oligofermentans glucose, ribose and xylose transcriptomes show higher similarity between glucose and xylose catabolism- induced responses in the early exponential growth phase. BMC Genomics 2016;17:539.

Battistuzzi FU, Feijao A, Hedges SB et al. A genomic timescale of prokaryote evolution: insights into the origin of methanogenesis, phototrophy, and the colonization of land. BMC Evol

Biol 2004;4:44.

van Bokhorst-van de Veen H, Smelt MJ, Wels M et al. Genotypic adaptations associated with prolonged persistence of Lactobacillus plantarum in the murine digestive tract. Biotechnol J

2013;8:895–904.

Bringel F, Curk MC, Hubert JC. Characterization of lactobacilli by southern-type hybridization with a Lactobacillus plantarum pyrDFE probe. Int J Syst Bacteriol 1996;46:588–94.

Broadbent JR, Neeno-Eckwall EC, Stahl B et al. Analysis of the Lactobacillus casei supragenome and its influence in species evolution and lifestyle adaptation. BMC Genomics

2012;13:533.

Bron P a., van Baarlen P, Kleerebezem M. Emerging molecular insights into the interaction between probiotics and the host intestinal mucosa. Nat Rev Microbiol 2011;10:66–78.

Bron PA, Grangette C, Mercenier A et al. Identification of Lactobacillus plantarum genes that are induced in the gastrointestinal tract of mice. J Bacteriol 2004a;186:5721–9.

Bron PA, Marco M, Hoffer SM et al. Genetic characterization of the bile salt response in

Lactobacillus plantarum and analysis of responsive promoters in vitro and in situ in the gastrointestinal tract. J Bacteriol 2004b;186:7829–35.

Brooijmans RJW, de Vos WM, Hugenholtz J. Lactobacillus plantarum WCFS1 electron transport chains. Appl Environ Microbiol 2009;75:3580–5.

Buhnik-Rosenblau K, Matsko-Efimov V, Jung M et al. Indication for co-evolution of

Lactobacillus johnsonii with its hosts. BMC Microbiol 2012;12:149.

Cai H, Rodríguez BT, Zhang W et al. Genotypic and phenotypic characterization of

Lactobacillus casei strains isolated from different ecological niches suggests frequent recombination and niche specificity. Microbiology 2007;153:2655–65.

Cai H, Thompson R, Budinich MF et al. Genome sequence and comparative genome analysis of Lactobacillus casei: insights into their niche-associated evolution. Genome Biol Evol

2009;1:239–57.

Ceapa C, Davids M, Ritari J et al. The Variable Regions of Lactobacillus rhamnosus Genomes

Reveal the Dynamic Evolution of Metabolic and Host-Adaptation Repertoires. Genome Biol Evol

2016;8:1889–905.

Chaillou S, Champomier-Vergès M-C, Cornet M et al. The complete genome sequence of the meat-borne lactic acid bacterium Lactobacillus sakei 23K. Nat Biotechnol 2005;23:1527–33.

Chaillou S, Lucquin I, Najjari A et al. Population genetics of Lactobacillus sakei reveals three lineages with distinct evolutionary histories. PLoS One 2013;8, DOI:

10.1371/journal.pone.0073253.

Chandler JA, Lang J, Bhatnagar S et al. Bacterial communities of diverse Drosophila species:

Ecological context of a host-microbe model system. PLoS Genet 2011;7:e1002272.

Clarke JT, Warnock RCM, Donoghue PCJ. Establishing a time-scale for plant evolution. New

Phytol 2011;192:266–301.

Daeschel MAA, Andersson REE, Fleming HPP. Microbial ecology of fermenting plant materials. FEMS Microbiol Lett 1987;46:357–67.

Dal Bello F, Hertel C. Oral cavity as natural reservoir for intestinal lactobacilli. Syst Appl

Microbiol 2006;29:69–76.

Danner H, Holzer M, Mayrhuber E et al. Acetic acid increases stability of silage under aerobic conditions. Appl Environ Microbiol 2003;69:562–7.

Delgado S, Suárez A, Mayo B. Dominant cultivable Lactobacillus species from the feces of healthy adults in northern Spain. Int Microbiol 2010;10:141–5.

Dewar WA, McDonald P, Whittenbury R. The hydrolysis of grass hemicelluloses during ensilage. J Sci Food Agric 1963;14:411–7.

Diancourt L, Passet V, Chervaux C et al. Multilocus sequence typing of Lactobacillus casei reveals a clonal population structure with low levels of homologous recombination. Appl Environ

Microbiol 2007;73:6601–11.

Dieckmann U, O’Hara B, Weisser W. The evolutionary ecology of dispersal. Trends Ecol Evol

1999;14:88–90.

Ding XF, Wu C De, Zhang LQ et al. Characterization of eubacterial and archaeal community diversity in the pit mud of Chinese Luzhou-flavor liquor by nested PCR-DGGE. World J Microbiol

Biotechnol 2014;30:605–12.

Douillard FP, Ribbera A, Järvinen HM et al. Comparative Genomic and Functional Analysis of

Lactobacillus casei and Lactobacillus rhamnosus Strains Marketed as Probiotics. Appl Environ

Microbiol 2013;79:1923–33.

Duar RM, Frese SA, Fernando SC et al. Experimental determination of host adaptation of

Lactobacillus reuteri to different vertebrate species. Appl Environ Microbiol 2017:AEM.00132-

17.(Chapter two of this thesis)

Ellegaard KM, Tamarit D, Javelind E et al. Extensive intra-phylotype diversity in lactobacilli and bifidobacteria from the honeybee gut. BMC Genomics 2015;16:284.

Endo A, Futagawa-Endo Y, Dicks LMT. Isolation and characterization of fructophilic lactic acid bacteria from fructose-rich niches. Syst Appl Microbiol 2009;32:593–600.

Endo A, Futagawa-Endo Y, Sakamoto M et al. Lactobacillus florum sp. nov., a fructophilic species isolated from flowers. Int J Syst Evol Microbiol 2010;60:2478–82.

Endo A, Irisawa T, Futagawa-Endo Y et al. Characterization and emended description of

Lactobacillus kunkeei as a fructophilic lactic acid bacterium. Int J Syst Evol Microbiol

2012;62:500–4.

Engel P, Moran NA. The gut microbiota of insects – diversity in structure and function. FEMS

Microbiol Rev 2013;37:699–735.

Erkosar B, Storelli G, Defaye A et al. Host-intestinal microbiota mutualism: “learning on the fly.” Cell Host Microbe 2013;13:8–14.

Fenton MP. An investigation into the sources of lactic acid bacteria in grass silage. J Appl

Bacteriol 1987;62:181–8.

Filannino P, Di Cagno R, Addante R et al. Metabolism of fructophilic lactic acid bacteria isolated from Apis mellifera L. bee-gut: a focus on the phenolic acids as external electron acceptors. Appl Environ Microbiol 2016, DOI: 10.1128/AEM.02194-16.

Frese SA, Benson AK, Tannock GW et al. The evolution of host specialization in the vertebrate gut symbiont Lactobacillus reuteri. PLoS Genet 2011;7:16.

Frese SA, Mackenzie DA, Peterson DA et al. Molecular characterization of host-specific biofilm formation in a vertebrate gut symbiont. PLoS Genet 2013;9:e1004057.

Fuller R, Brooker BE. Lactobacilli which attach to the crop epithelium of the fowl. Am J Clin

Nutr 1974;27:1305–12.

Galdeano CM, Perdigón G. Role of viability of probiotic strains in their persistence in the gut and in mucosal immune stimulation. J Appl Microbiol 2004;97:673–81.

Gänzle M, Ripari V. Composition and function of sourdough microbiota: From ecological theory to bread quality. Int J Food Microbiol 2016;239:19–25.

Gänzle MG. Lactic metabolism revisited: metabolism of lactic acid bacteria in food fermentations and food spoilage. Curr Opin Food Sci 2015;2:106–17.

Gänzle MG, Follador R. Metabolism of Oligosaccharides and Starch in Lactobacilli: A Review.

Front Microbiol 2012;3:340.

Gänzle MG, Vermeulen N, Vogel RF. Carbohydrate, peptide and lipid metabolism of lactic acid bacteria in sourdough. Food Microbiol 2007;24:128–38.

Grill JP, Cayuela C, Antoine JM et al. Isolation and characterization of a Lactobacillus amylovorus mutant depleted in conjugated bile salt hydrolase activity: relation between activity and bile salt resistance. J Appl Microbiol 2000;89:553–63.

Guan LL, Hagen KE, Tannock GW et al. Detection and identification of Lactobacillus species in crops of broilers of different ages by using PCR-denaturing gradient gel electrophoresis and amplified ribosomal DNA restriction analysis. Appl Environ Microbiol 2003;69:6750–7.

van de Guchte M, Penaud S, Grimaldi C et al. The complete genome sequence of

Lactobacillus bulgaricus reveals extensive and ongoing reductive evolution. Proc Natl Acad Sci

U S A 2006;103:9274–9.

Hallmann J, Quadt-Hallmann A, Mahaffee WF et al. Bacterial endophytes in agricultural crops.

Can J Microbiol 1997;43:895–914.

Hammes WP, Hertel C. The Genera Lactobacillus and Carnobacterium. The Prokaryotes.

New York, NY, NY: Springer US, 2006, 320–403.

Hammons S, Oh PL, Martínez I et al. A small variation in diet influences the Lactobacillus strain composition in the crop of broiler chickens. Syst Appl Microbiol 2010;33:275–81.

Hayden B, Canuel N, Shanse J. What was brewing in the natufian? An archaeological assessment of brewing technology in the epipaleolithic. J Archaeol Method Theory 2013;20:102–

50.

Hedges SB, Blair JE, Venturi ML et al. A molecular timescale of eukaryote evolution and the rise of complex multicellular life. BMC Evol Biol 2004;4:2.

Heilig HGHJ, Zoetendal EG, Vaughan EE et al. Molecular diversity of lactobacillus spp. and other lactic acid bacteria in the human intestine as determined by specific amplification of 16S ribosomal DNA. Appl Environ Microbiol 2002;68:114–23.

Hill C, Guarner F, Reid G et al. Expert consensus document: The International Scientific

Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat Rev Gastroenterol Hepatol 2014;advance on, DOI:

10.1038/nrgastro.2014.66.

Jacobs MJ, Bugbee WM, Gabrielson DA. Enumeration, location, and characterization of endophytic bacteria within sugar beet roots. Can J Bot 1985;63:1262–5.

Kakimoto S, Okazaki K, Sakane T et al. Isolation and Taxonomic Characterization of Acid

Urease-Producing Bacteria. Agric Biol Chem 1989;53:1111–8.

Kandler O WN. Genus Lactobacillus. In: Sneath P, Mair N, Sharpe M, et al. (eds.). Bergey’s

Manual of Systematic Bacteriology. 9th ed. Williams & Wilkins, 1986, 1063–1065.

Kankainen M, Paulin L, Tynkkynen S et al. Comparative genomic analysis of Lactobacillus rhamnosus GG reveals pili containing a human- mucus binding protein. Proc Natl Acad Sci U S

A 2009;106:17193–8.

Kitara K, Kaneko T, Goto O. Taxonomic studies on the hiochi-bacteria, specific saprophytes of sake. II. Identification and classification of hiochi-bacteria. J Gen Appl Microbiol 1957;3:111–

20.

Krumbeck JA, Marsteller NL, Frese SA et al. Characterization of the ecological role of genes mediating acid resistance in Lactobacillus reuteri during colonization of the gastrointestinal tract.

Environ Microbiol 2016;18:2172–84.

Kvasnikov EI, Kovalenko NK, Nesterenko OA. Lactic acid bacteria in nature and the national economy. Appl Biochem Microbiol 1983:665–76.

Kwong WK, Moran NA. Gut microbial communities of social bees. Nat Rev Microbiol

2016;14:374–84.

Lähteinen T, Malinen E, Koort JMK et al. Probiotic properties of Lactobacillus isolates originating from porcine intestine and feces. Anaerobe 2010;16:293–300.

De Las Rivas B, Marcobal ÁE, Gómez A et al. Characterization of ISLpl4, a functional insertion sequence in Lactobacillus plantarum. Gene 2005;363:202–10.

Lebeer S, Vanderleyden J, De Keersmaecker SCJ. Genes and molecules of lactobacilli supporting probiotic action. Microbiol Mol Biol Rev 2008;72:728–764, Table of Contents.

Lo W-SS, Huang Y-YY, Kuo C-HH. Winding paths to simplicity: genome evolution in facultative insect symbionts. Lai E-M (ed.). FEMS Microbiol Rev 2016;40:855–74.

Luo Z-X. Transformation and diversification in early mammal evolution. Nature

2007;450:1011–9.

Mackenzie DA, McLay K, Roos S et al. Draft Genome sequence of a Novel Lactobacillus salivarius strain isolated from piglet. Genome Announc 2014;2, DOI: 10.1128/genomeA.01231-

13.

Macklaim JM, Gloor GB, Anukam KC et al. At the crossroads of vaginal health and disease, the genome sequence of Lactobacillus iners AB-1. Proc Natl Acad Sci U S A 2011:4688–95.

Maeno S, Tanizawa Y, Kanesaki Y et al. Genomic characterization of a fructophilic bee symbiont Lactobacillus kunkeei reveals its niche-specific adaptation. Syst Appl Microbiol

2016;39:516–26.

Makarova K, Slesarev a, Wolf Y et al. Comparative genomics of the lactic acid bacteria. Proc

Natl Acad Sci U S A 2006;103:15611–6.

Mann E, Schmitz-Esser S, Zebeli Q et al. Mucosa-Associated Bacterial Microbiome of the

Gastrointestinal Tract of Weaned Pigs and Dynamics Linked to Dietary Calcium-Phosphorus.

Blachier F (ed.). PLoS One 2014;9:e86950.

Marco ML, Pavan S, Kleerebezem M. Towards understanding molecular modes of probiotic action. Curr Opin Biotechnol 2006;17:204–10.

Martín R, Jiménez E, Olivares M et al. Lactobacillus salivarius CECT 5713, a potential probiotic strain isolated from infant feces and breast milk of a mother–child pair. Int J Food

Microbiol 2006;112:35–43.

Martínez I, Stegen JC, Maldonado-Gómez MX et al. The gut microbiota of rural papua new guineans: composition, diversity patterns, and ecological processes. Cell Rep 2015;11:527–38.

Martino ME, Bayjanov JR, Caffrey BE et al. Nomadic lifestyle of Lactobacillus plantarum revealed by comparative genomics of 54 strains isolated from different habitats. Environ Microbiol

2016;18:4974–89.

Matos RC, Leulier F. Lactobacilli-Host mutualism: “learning on the fly.” Microb Cell Fact

2014;13:S6.

McFrederick QS, Cannone JJ, Gutell RR et al. Specificity between lactobacilli and hymenopteran hosts is the exception rather than the rule. Appl Environ Microbiol 2013;79:1803–

12.

McFrederick QS, Wcislo WT, Taylor DR et al. Environment or kin: whence do bees obtain acidophilic bacteria? Mol Ecol 2012;21:1754–68.

de Melo Pereira GV, Magalhães KT, Lorenzetii ER et al. A multiphasic approach for the identification of endophytic bacterial in strawberry fruit and their potential for plant growth promotion. Microb Ecol 2012;63:405–17.

Mendes-Soares H, Suzuki H, Hickey RJ et al. Comparative functional genomics of

Lactobacillus spp. reveals possible mechanisms for specialization of vaginal lactobacilli to their environment. J Bacteriol 2014;196:1458–70.

Mercier J, Lindow SE. Role of leaf surface sugars in colonization of plants by bacterial epiphytes. Appl Environ Microbiol 2000;66:369–74.

Minervini F, Celano G, Lattanzi A et al. Lactic acid bacteria in durum wheat flour are endophytic components of the plant during its entire life cycle. Griffiths MW (ed.). Appl Environ

Microbiol 2015;81:6736–48.

Misof B, Liu S, Meusemann K et al. Phylogenomics resolves the timing and pattern of insect evolution. Science (80- ) 2014;346.

Molenaar D, Bringel F, Schuren FH et al. Exploring Lactobacillus plantarum genome diversity by using microarrays. J Bacteriol 2005;187:6119–27.

Molin G, Jeppsson B, Johansson M-L et al. Numerical taxonomy of Lactobacillus spp. associated with healthy and diseased mucosa of the human intestines. J Appl Bacteriol

1993;74:314–23.

Moran NA. Symbiosis. Curr Biol 2006;16:R866–71.

Müller M, Lier D. Fermentation of fructans by epiphytic lactic acid bacteria. J Appl Bacteriol

1994;76:406–11.

Mundt JO, Hammer JL. Lactobacilli on plants. Appl Microbiol 1968;16:1326–30.

Nakamura LK. Lactobacillus amylovorus, a New Starch-Hydrolyzing Species from Cattle

Waste-Corn Fermentations. Int J Syst Bacteriol 1981;31:56–63.

Neveling DP, Endo A, Dicks LMT. Fructophilic Lactobacillus kunkeei and Lactobacillus brevis isolated from fresh flowers, bees and bee-hives. Curr Microbiol 2012;65:507–15.

van den Nieuwboer M, van Hemert S, Claassen E et al. Lactobacillus Plantarum WCFS1 and

Its Host Interaction: A Dozen Years after the Genome. Wiley-Blackwell, 2016:452–65.

O’ Donnell MM, Harris HMB, Lynch DB et al. Lactobacillus ruminis strains cluster according to their mammalian gut source. BMC Microbiol 2015;15:80.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Pedersen K, Tannock GW. Colonization of the porcine gastrointestinal tract by lactobacilli.

Appl Envir Microbiol 1989;55:279–83.

Petrova MI, Reid G, Vaneechoutte M et al. Lactobacillus iners: Friend or Foe? Trends

Microbiol 2016;xx:1–10.

Pires ND, Dolan L. Morphological evolution in land plants: new designs with old genes. Philos

Trans R Soc B Biol Sci 2012;367:508–18.

Powell JE, Leonard SP, Kwong WK et al. Genome-wide screen identifies host colonization determinants in a bacterial gut symbiont. Proc Natl Acad Sci U S A 2016;113:13887–92.

Raftis EJ, Salvetti E, Torriani S et al. Genomic Diversity of Lactobacillus salivarius. Appl

Environ Microbiol 2011;77:954–65.

Regmi PR, Metzler-Zebeli BU, Ganzle MG et al. Starch with high amylose content and low in vitro digestibility increases intestinal nutrient flow and microbial fermentation and selectively promotes bifidobacteria in pigs. J Nutr 2011;141:1273–80.

Reti KL, Thomas MC, Yanke LJ et al. Effect of antimicrobial growth promoter administration on the intestinal microbiota of beef cattle. Gut Pathog 2013;5:8.

Reuter G. The Lactobacillus and Bifidobacterium microflora of the human intestine: composition and succession. Curr Issues Intest Microbiol 2001;2:43–53.

Ripari V, Gänzle MG, Berardi E. Evolution of sourdough microbiota in spontaneous sourdoughs started with different plant materials. Int J Food Microbiol 2016;232:35–42.

Rogosa M, Wiseman RF, Mitchell JA et al. Species differentiation of oral lactobacilli from man including description of Lactobacillus salivarius nov spec and Lactobacillus Cellobiosus nov spec.

J Bacteriol 1953;65:681–99.

Rossi M, Martínez-Martínez D, Amaretti A et al. Mining metagenomic whole genome sequences revealed subdominant but constant Lactobacillus population in the human gut microbiota. Environ Microbiol 2016:n/a-n/a.

Ruiz L, Margolles A, Sánchez B. Bile resistance mechanisms in Lactobacillus and

Bifidobacterium. Front Microbiol 2013;4, DOI: 10.3389/fmicb.2013.00396.

Sachs J, Skophammer R, Regus J. Evolutioanry transitions in bacterial symbiosis. PNAS

2011;108:10800–7.

Scheirlinck I, Van der Meulen R, De Vuyst L et al. Molecular source tracking of predominant lactic acid bacteria in traditional Belgian sourdoughs and their production environments. J Appl

Microbiol 2009;106:1081–92.

Shetty S, Griffin DK, Graves JAM. Comparative Painting Reveals Strong Chromosome

Homology Over 80 Million Years of Bird Evolution. Chromosom Res 1999;7:289–95.

Shrivastava KPP. Bacterium (Lactobacillus?) eurydice strains from bumble bees. J Invertebr

Pathol 1982;40:180–5.

Siezen RJ, van Hylckama Vlieg JET. Genomic diversity and versatility of Lactobacillus plantarum, a natural metabolic engineer. Microb Cell Fact 2011;10 Suppl 1:S3.

Siezen RJ, Tzeneva VA, Castioni A et al. Phenotypic and genomic diversity of Lactobacillus plantarum strains isolated from various environmental niches. Environ Microbiol 2010;12:758–73.

Smokvina T, Wels M, Polka J et al. Lactobacillus paracasei Comparative Genomics: Towards

Species Pan-Genome Definition and Exploitation of Diversity. Highlander SK (ed.). PLoS One

2013;8:e68731.

Stefanovic E, Fitzgerald G, McAuliffe O. Advances in the genomics and metabolomics of dairy lactobacilli: A review. Food Microbiol 2017;61:33–49.

Steinkraus KH. Fermentations in World Food Processing. Compr Rev Food Sci Food Saf

2002;1:23–32.

Stiles ME, Holzapfel WH. Lactic acid bacteria of foods and their current taxonomy. Int J Food

Microbiol 1997;36:1–29.

Stirling AC, Whittenbury R. Sources of the Lactic Acid Bacteria Occurring in Silage. J Appl

Bacteriol 1963;26:86–90.

Su MS-W, Oh PL, Walter J et al. Intestinal origin of sourdough Lactobacillus reuteri isolates as revealed by phylogenetic, genetic, and physiological analysis. Appl Environ Microbiol

2012;78:6777–80.

Sun Z, Harris HMB, McCann A et al. Expanding the biotechnology potential of lactobacilli through comparative genomics of 213 strains and associated genera. Nat Commun 2015;6:8322.

Tailliez P. Mini-revue : les bactéries lactiques,ces êtres vivants apparus il y a près de 3 milliards d’années. Lait 2001;81:1–11.

Tamarit D, Ellegaard KM, Wikander J et al. functionally structured genomes in Lactobacillus kunkeei colonizing the honey crop and food products of honeybees and stingless bees. Genome

Biol Evol 2015;7:1455–73.

Tannock GW. The Lactic Microflora of Pigs, Mice and Rats. The Lactic Acid Bacteria Volume

1. Boston, MA, MA: Springer US, 1992, 21–48.

Tannock GW. A special fondness for lactobacilli. Appl Environ Microbiol 2004;70:3189–94.

Tannock GW, Munro K, Harmsen HJ et al. Analysis of the fecal microflora of human subjects consuming a probiotic product containing Lactobacillus rhamnosus DR20. Appl Environ Microbiol

2000;66:2578–88.

Tannock GW, Wilson CM, Loach D et al. Resource partitioning in relation to cohabitation of

Lactobacillus species in the mouse forestomach. ISME J 2012;6:927–38.

Techo S, Miyashita M, Shibata C et al. Lactobacillus ixorae sp. nov., isolated from a flower

(West-Indian jasmine). Int J Syst Evol Microbiol 2016;66:5500–5.

Torriani S, Clementi F, Vancanneyt M et al. Differentiation of Lactobacillus plantarum, L. pentosus and L. paraplantarum Species by RAPD-PCR and AFLP. Syst Appl Microbiol

2001;24:554–60.

Vásquez A, Forsgren E, Fries I et al. Symbionts as major modulators of insect health: lactic acid bacteria and honeybees. PLoS One 2012;7:e33188.

Vellend M. Conceptual synthesis in community ecology. Q Rev Biol 2010;85:183–206.

Vera Pingitore E, Hébert EM, Nader-Macías ME et al. Characterization of salivaricin CRL

1328, a two-peptide bacteriocin produced by Lactobacillus salivarius CRL 1328 isolated from the human vagina. Res Microbiol 2009;160:401–8.

Vogel RF, Pavlovic M, Ehrmann MA et al. Genomic analysis reveals Lactobacillus sanfranciscensis as stable element in traditional sourdoughs. Microb Cell Fact 2011;10 Suppl

1:S6.

Vojvodic S, Rehan SM, Anderson KKE et al. Microbial Gut Diversity of Africanized and

European Honey Bee Larval Instars. PLoS One 2013;8:e72106.

Wall R, Fitzgerald G, Hussey S et al. Genomic diversity of cultivable Lactobacillus populations residing in the neonatal and adult gastrointestinal tract. FEMS Microbiol Ecol 2007;59:127–37.

Walter J. Ecological role of lactobacilli in the gastrointestinal tract: implications for fundamental and biomedical research. Appl Environ Microbiol 2008;74:4985–96.

Walter J, Britton RA, Roos S. Host-microbial symbiosis in the vertebrate gastrointestinal tract and the Lactobacillus reuteri paradigm. Proc Natl Acad Sci U S A 2011;108 Suppl:4645–52.

Walter J, Chagnaud P, Tannock GW et al. A high-molecular-mass surface protein (Lsp) and methionine sulfoxide reductase B (MsrB) contribute to the ecological performance of Lactobacillus reuteri in the murine gut. Appl Environ Microbiol 2005;71:979–86.

Walter J, Hertel C, Tannock GW et al. Detection of Lactobacillus, Pediococcus, Leuconostoc, and Weissella species in human feces by using group-specific pcr primers and denaturing gradient gel electrophoresis. Appl Environ Microbiol 2001;67:2578–85.

Walter J, Loach DM, Alqumber M et al. D-alanyl ester depletion of teichoic acids in

Lactobacillus reuteri 100-23 results in impaired colonization of the mouse gastrointestinal tract.

Environ Microbiol 2007;9:1750–60.

Wang C-Y, Lin P-R, Ng C-C et al. Probiotic properties of Lactobacillus strains isolated from the feces of breast-fed infants and Taiwanese pickled cabbage. Anaerobe 2010;16:578–85.

Wegmann U, MacKenzie DA, Zheng J et al. The pan-genome of Lactobacillus reuteri strains originating from the pig gastrointestinal tract. BMC Genomics 2015;16:1023.

Wong AC-N, Chaston JM, Douglas AE. The inconstant gut microbiota of Drosophila species revealed by 16S rRNA gene analysis. ISME J 2013;7:1922–32.

Wong ACN, Luo Y, Jing X et al. The host as the driver of the microbiota in the gut and external environment of Drosophila melanogaster. Appl Environ Microbiol 2015;81:6232–40.

Wong CNA, Ng P, Douglas AE. Low-diversity bacterial community in the gut of the fruitfly

Drosophila melanogaster. Environ Microbiol 2011;13:1889–900.

Yuki N, Shimazaki T, Kushiro A et al. Colonization of the stratified squamous epithelium of the nonsecreting area of horse stomach by lactobacilli. Appl Environ Microbiol 2000;66:5030–4.

Zheng J, Ruan L, Sun M et al. A Genomic View of Lactobacilli and Pediococci Demonstrates that Phylogeny Matches Ecology and Physiology. Björkroth J (ed.). Appl Environ Microbiol

2015a;81:7233–43.

Zheng J, Zhao X, Lin XB et al. Comparative genomics Lactobacillus reuteri from sourdough reveals adaptation of an intestinal symbiont to food fermentations. Sci Rep 2015b;5:18234.

Zotta T, Ricciardi A, Parente E et al. Draft Genome Sequence of the Respiration-Competent

Strain Lactobacillus casei N87. Genome Announc 2016;4:e00348-16.

2. Chapter two: Experimental evaluation of host adaptation of Lactobacillus reuteri to different vertebrate species

A version of Chapter two of this thesis was published as: Duar RM, Frese SA, Fernando SC et al. Experimental determination of host adaptation of Lactobacillus reuteri to different vertebrate species. Appl Environ Microbiol 2017:AEM.00132-17.

2.1 Introduction

Vertebrates have gained access to a number of metabolic functions absent in their genomes by virtue of symbiosis with microbes (Moran 2006). The crop and the cecum are evident anatomical adaptations to exploit the enzymatic activity of trillions of symbiotic microbes to access nutrients from otherwise indigestible diets (Stevens and Huma 1998). In addition to nutrient provision, gut microbes exclude pathogens and aid in development of the host’s immune system

(Dethlefsen, McFall-Ngai and Relman 2007). It is clear vertebrates benefit from symbiotic associations with microbes. However, core concepts on how symbioses are formed and maintained over evolutionary time scales are not well understood.

The taxonomic profile of the vertebrate microbiota is largely host-specific (Dethlefsen, McFall-

Ngai and Relman 2007; Ley et al. 2008a) and in some cases, congruent with the evolution of the host species (Ochman et al. 2010). This apparent relatedness between microbial community composition and host phylogeny has been interpreted as evidence for co-evolution (Ley et al.

2008b; Fraune and Bosch 2010; Rosenberg and Zilber-Rosenberg 2016). Much of this information has been derived from analyzing 16S rRNA gene sequences which have evolved too slowly to provide information on evolutionary relationships over relevant time scales, especially given the more recent diversification of contemporary vertebrate species compared to their bacterial symbionts (Walter, Britton and Roos 2011). Examples of co-diversification of specific bacterial lineages with mammalian hosts have been discovered by analyzing fast-evolving gene phylogenies (Falush et al. 2003; Moeller et al. 2016) suggesting co-speciation of some symbiotic microbes alongside their vertebrate hosts. However, even in the few established cases, the mechanisms by which these microbes evolved and the outcomes that arose from the evolutionary process remain undefined.

Symbiotic gut microbes that remain stably associated with particular vertebrate species are predicted to evolve host-specific adaptations and as a result, display enhanced ecological performance in their cognate host (Garcia and Gerardo 2014). Determining the rate of colonization success of individual symbionts in naïve (aposymbiotic) hosts in combination with phylogenetic analyses thus provides a platform to infer the evolutionary mechanisms by which host-microbe symbioses evolve. Such experimental approach has been successfully applied to study the evolutionary consequences of bacterial symbioses in insects (Dale and Moran 2006;

Kwong et al. 2014) and in invertebrates (Wollenberg and Ruby 2012; Bongrand et al. 2016), but rarely in vertebrates.

The bacterial species Lactobacillus reuteri inhabits the gastrointestinal tracts of a variety of vertebrates and has diversified into distinct phylogenetic lineages that are coherent with host origin (Oh et al. 2010). A series of phylogenetic, phylogenomic and experimental studies in mice have established this species as a paradigm for host adaptation of a non-pathogenic symbiont of the vertebrate gut microbiota (Walter, Britton and Roos 2011). Rodent isolated strains display elevated fitness in mice (Oh et al. 2010; Frese et al. 2011) and biofilm formation in the forestomach is restricted to strains from rodent lineages (Frese et al. 2013). Together these results show that specialized adaptations to the gut environment of the host underlie the evolution of L. reuteri with rodents. In contrast, the mechanisms by which L. reuteri has evolved with other vertebrate host species have not been determined. In this respect, it is important to point out that the presence of host-specific lineages in itself does not provide evidence for natural selection as clusters can arise by neutral processes such as genetic drift (Doolittle and Zhaxybayeva 2009;

Oh et al. 2010). Furthermore, clustering of both human and poultry isolates in the same phylogenetic lineage (lineage-VI) has raised questions regarding the evolutionary history of this lineage. Therefore, the goals of this study were to test for host adaptation of L. reuteri to non- rodent hosts and to resolve outstanding questions regarding the evolution of the phylogenetic

lineage-VI. We developed and validated an experimental method to systematically compare the ecological performance of strains isolated from different hosts and assigned to distinct phylogenetic lineages, in the gastrointestinal tracts of chicken, pigs and human volunteers. We then complemented these studies with comparative genomic analyses to gain insight into genome evolution of poultry and human strains of the lineage-VI.

2.2 Materials and Methods

2.2.1 Strains, media and growth conditions

Bacterial strains used in this study are listed in Table 2.1 and were grown at 37 °C in deMan

Rogosa Sharpe medium supplemented with 10 g/l maltose and 5 g/l of fructose (mMRS). To ensure selective cultivation from feces, a L. reuteri isolation medium (LRIM) was devised based on the recipe of the Rogosa medium (Rogosa, Mitchell and Wiseman 1951). The LRIM contained per liter, 15 g of raffinose, 15 g of sodium acetate, 15 g of agar, 10 g of tryptone, 6 g of KH2PO4,

5 g of yeast extract, 2 g of ammonium acetate, 1.32 ml of glacial acetic acid, 1 g of tween 80, 0.57 g of MgSO4, of 0.12 g of MnSO4, 0.003 g of FeSO4. Since only a limited number of lactobacilli species grow on raffinose (Pot et al. 2014), this media allowed for a sufficiently selective culture of L. reuteri in the presence of a background fecal microbiota. Incubations in LRIM were performed under anaerobic conditions (<0.1 % O2 ; ≥15% CO2) for 48 h at 45 °C.

2.2.2 Preparation of strain mixtures to prepare inocula

Eighteen L. reuteri strains originating from different hosts and assigned to separate phylogenetic lineages were selected and divided into three mixtures (see Table 2.1). To facilitate differentiation, strains within the same inoculum were selected to carry distinct leuS alleles.

Inocula for host experiments were prepared by growing individual stains overnight in mMRS, followed by subculture (with 1% inoculum each) twice in previously boiled (100 °C for 30 min) food-grade DE-PHAGE® media (Cargill) supplemented with 20 g/l of malt. Cell numbers of each

individual strain after growth for 16 h in DE-PHAGE® media were determined in at least six replicate experiments. This information was used to prepare standardized inocula with equivalent proportions of each strain and adjusted to contain approximately 3 x 105 cells of total L. reuteri per gram of host body weight (BW). In order to determine if these conditions were met, cell numbers of each individual stain were determined by quantitative culture on mMRS, prior to mixing the inoculums

Table 2.1- L. reuteri strains used in the host adaptation assay Strain Host of Lineagea leuS allelic isolation profilea Inoculum A Cf46g Human II 4 mlc3 Mouse III 31 CR Rat I 14 CSF8 Chicken VI 15 M27U15 Human VI 11 JW2015 Pig IV 3

Inoculum B MM4-1a (ATCCPTA 6475) Human II 4 r2lc Rat III 35 lpuph1 Mouse I 9 1366 Chicken VI 6 CF48-3A Human VI 11 lpa1 Pig IV 3

Inoculum C sr11 Human II 4 n2d Rat III 34 6799jm1 Mouse I 9 JCM1081 Chicken VI 24 MM34-4A Human VI 11 ATCC 53608 Pig IV 3 a Lineage and leuS type as determined in (Oh et al. 2010)

2.2.3 Mouse experiment

Germ-free C3H/HeN mice (~30 g BW) were maintained at the University of Nebraska

Gnotobiotic Mouse Facility. Groups of mice (n=5) were assigned to receive one of the three inocula and subsequently moved from sterile isolators into individual ventilated biocontainment cages. To prepare the inocula, L. reuteri strains grown for 16 h on DE-PHAGE® media were harvested by centrifugation (3000  g for 10 min) and washed twice with sterile PBS (pH 7.0).

Each mouse was gavaged with 100 µl of PBS suspension containing a total of 3 x 105 cells per gram of BW. L. reuteri were enumerated on LRIM from fresh fecal pellets collected immediately prior and daily for 5 days after gavage. To ensure gnotobiotic conditions were maintained during the experiments, three mice were gavaged with sterile PBS and housed in separate biocontainment cages located in the same ventilator rack as the experimental mice. Fecal pellets from control mice were plated daily on BHI and mMRS media and checked for aerobic and anaerobic growth. All procedures were conducted with approval from the Institutional Animal Care and Use Committee of the University of Nebraska (Project ID 731).

2.2.4 Chicken experiment

Specific pathogen free Leghorns were hatched in the Poultry Research Facility at the

University of Alberta and transported the same day to the Animal Research Facility of the same institution. Birds were randomly assigned to an inoculum (n=5 per inoculum) or PBS control (4 birds). Chickens were housed in pairs or groups of three and maintained in biocontainment cages.

In order to obtain Lactobacillus-free (LF) chickens penicillin was added to the drinking water at a concentration of 0.6 g/l. Four days after days of penicillin treatment commenced, absence of lactobacilli was confirmed by plating cloacal swabs on Rogosa (Difco) and LRIM. Antibiotic administration was removed 18 h prior to LF-chickens (~ 30g BW) being gavaged with 300 µl of

PBS containing standardized L. reuteri inocula with a total of 3  105 cells per gram of BW. Cloacal

samples were obtained from each animal immediately prior (day 0) and for five days after gavage.

Upon collection, cloacal samples were transferred into 1.5 ml tubes containing sterile PBS with

10% v/v glycerol and immediately processed by dilution plating on LRIM. All procedures were carried out in accordance with protocol AUP00000003 approved by the University of Alberta’s

Animal Care and Use Committee.

2.2.5 Pig experiment

Nine germ-free piglets were delivered from a full-term pregnant sow by sterile hysterectomy following methods described by Miniats and Jol (Miniats and Jol 1978) and aseptically transferred into one of three sterile polyvinyl flexible isolators. Animals in the same isolator (n=3) were kept in separate stainless steel compartments with false floors to collect excreta. Isolators were maintained under positive pressure at an ambient temperature of 35°C and ventilated with sterilized filtered air pre-warmed to the same temperature. Piglets were fed commercially sterile infant formula throughout the experiment. At 10 days of age, piglets (~3 Kg BW) in the same isolator were administered either inocula A, B or C containing L. reuteri at 3  105 cells per gram of BW and suspended in the feeding infant formula. Cell counts of L. reuteri were enumerated by quantitative LRIM culture of fecal samples obtained directly from the pig’s rectum using sterile plastic loops and collected before (day 0) and at days 1 to 5 post-administration. Fecal samples were also plated on BHI to detect any contamination. Aerobic growth on BHI was detected in the feces of one of the treatment groups on day 3 to 5 post-inoculation. However, L. reuteri counts were not different when compared with pigs in the other groups. All procedures were conducted in with approval of the Institutional Animal Care and Use Committee of the University of Nebraska

(Project ID 939).

2.2.6 Human subjects

Fecal samples from twenty human subjects were screened for growth on LRIM. Of those, fifteen (eight females, seven males) produced less than 104 CFU per gram of feces and were considered eligible for the study. Subjects were then randomly assigned to receive one of the three inocula (n=5 per inoculum). All subjects were between the ages of 18 and 55, abstained from using probiotic and prebiotic products, had not consumed oral antibiotics within 3 months before the study and considered themselves healthy.

Inocula were prepared in the food laboratory of the Department of Agricultural, Food, and

Nutritional Science at the University of Alberta. Briefly, L. reuteri strains were grown for 16 h on

DE-PHAGE® media and mixed to contain a total of 3  105 cells per gram of BW (considering an average BW of 75 kg). Cells were harvested by centrifugation at 3000  g for 5 min and suspended in bottled spring water immediately prior to consumption. Subjects were instructed to drink contents of the solution in a single setting. Fecal samples were collected daily from day 0 (before consumption) and during days 1 to 5 post-consumption. Samples were processed within 2 h of deposition and L. reuteri was cultured by quantitative dilution plating on LRIM. Human studies were completed at the University of Alberta in accordance to the protocol Pro00051493 approved by The Health Research Ethics Board - Biomedical panel.

2.2.7 Strain typing and identification.

Sixteen colonies were randomly selected from LRIM plates of each fecal sample (or cloacal swab) cultured from day 1 to day 5 post-inoculation and typed by colony PCR. The leuS gene was directly amplified from each colony with the primers leuS-F TACGACGCGGGCAGATAC and leuS-R ATAGAGATCAACTGGTGACC. PCR conditions described previously (Oh et al. 2010) were as follows: 94°C for 2 min, followed by 30 cycles of 94°C for 30s, 55°C for 30s, and 72°C for 1 min, with a final extension of 7 min at 72°C. PCR products were purified using the QIAquick

PCR Purification Kit (Qiagen) and sequenced using Sanger technology. Sequences were assigned to a strain by BLASTn search against a local nucleotide database implemented in

Bioedit (Hall 1999).

2.2.8 Genome sequencing and annotation

Genomic DNA from the strains L. reuteri 1366, CSF8, JCM1081 (poultry-VI), and MM34-4A,

M27U15 and CF48-3A1 (human-VI) was obtained using the Qiagen DNeasy Blood and Tissue

Kit (Qiagen) with some modifications as described by Oh et. al (2010) (Oh et al. 2010). Genomes were sequenced to draft status at The Applied Genomics Centre (TAGC, University of Alberta) using Illumina MiSeq paired-end technology. Reads were assembled into scaffolds with SPADES

(Bankevich et al. 2012) available in PATRIC (Wattam et al. 2014), annotated using the Joint

Genome Institute (http://jgi.doe.gov) pipeline and deposited in the Integrated Microbial Genomes system (IMG) (Markowitz et al. 2008).

2.2.9 Accession numbers

The genome sequences of L. reuteri sequenced in this study were deposited in the Integrated

Microbial Genomes system (IMG) (Markowitz et al. 2008). Genome IDs are provided in Table 2.

Genome sequences are also available in the GenBank under the BioProject ID: PRJNA380292.

2.2.10 Comparative genomic and phylogenetic analyses

Comparative analysis of genome sequences was performed in EDGAR (Blom et al. 2016) based on an all-against-all comparison of the predicted proteomes (GenBank) downloaded from the JGI. Unique and lineage-specific genes were determined using the “singleton” and “genesets” functions in EDGAR (Blom et al. 2016) and confirmed with the IMG phylogenetic profiler

(Markowitz et al. 2008). Single nucleotide polymorphisms were detected with Mauve (Darling,

Mau and Perna 2010) and the average nucleotide identity (ANI) in EDGAR (Blom et al. 2016).

The core genome was calculated as the set of orthologous genes present in all strains by bidirectional best BLAST hits. Phylogeny was constructed by aligning 900 core orthologous genes present in all the genomes sequenced in this study and others available in the public databases

(Table 1). Concatenated sequences were used to calculate a distance matrix, which provided the input for the neighbor-joining method in the PHYLIP package as implemented in EDGAR (Blom et al. 2016). Tree was drawn and annotated using Figtree

(http://tree.bio.ed.ac.uk/software/figtree/)

2.2.11 Ancestral state analysis

To infer the order of emergence of host lineages, the sequences of the seven gene of 116 strains used in previous multi-locus sequence analysis (Oh et al. 2010) were analyzed using

Mesquite 3.2 (Maddison and Maddisson 2017) as previously described (Sachs, Skophammer and

Regus 2011). Lactobacillus fermentum was used as an outlier. The host of each strain was assigned as rodent, swine, poultry or human based on the origin of isolation (Oh et al. 2010).

Ancestral states were inferred using parsimony. Host switch events were identified when two or more equally parsimonious ancestral state reconstructions were found, or when the host state of the immediate ancestral node was different from the offspring. The putative dates of host switch events were estimated using Bayesian phylogenetic analyses in BEAST v.2.4.3 (Bouckaert et al.

2014) using the HKY85 substitution model, an estimated clock rate of 10-8 and the Calibrated

Yule model.

2.2.12 Statistical analysis

Statistical significance was determined by one-way analysis of variances (ANOVA) with

Tukey’s post hoc test (α = 0.05) implemented in the statistical package GraphPad Prism version

6.0 (GraphPad Software, La Joya, CA, USA).

2.3 Results

2.3.1 Evolutionary relationships of L. reuteri strains using whole genome phylogenetic

analysis

To achieve a higher resolution in the analysis of the evolutionary relationships of strains

belonging to lineage-VI, we sequenced the genomes of six strains from this cluster (3 originating

from chickens and 3 from humans) and included these genomes in a whole-genome phylogenetic

Figure 2.1 Neighbor-joining tree of Lactobacillus reuteri Phylogenetic tree of Lactobacillus reuteri based on core genome alignment (900 genes) of 25 strains. Tips of the branches are color coded by lineage and cohesive clades are labeled.

analysis with an additional nineteen L. reuteri genomes (Table 2). As shown in Fig. 2.1, the tree showed clear separation of L. reuteri into six host-defined phylogenetic lineages in complete agreement with the previous multi-locus sequencing analysis (Oh et al. 2010). Whole-genome phylogenetic analysis further confirmed human and poultry isolates clustered in lineage VI, which we refer to as human-VI and poultry-VI strains, respectively. This analysis also revealed important differences in genome diversity among lineage-VI strains, with the genomes of human-VI clustered tightly while poultry-VI strains showed high diversity (Fig. 2.1).

2.3.2 Introduction of L. reuteri to the digestive tract of different vertebrate hosts.

We developed an experimental approach to study host adaptation of L. reuteri strains in different hosts (Fig. 2.2). We designed three inocula each containing six L. reuteri strains representing lineage I, II, III, V and VI, with isolates originating from rodents (mice or rats), pigs, chicken, and humans, including one chicken-VI strain and one human-VI strain in each inoculum

(Table 2.1). We administrated the same three inocula at a standardized dose to germ-free mice and germ-free pigs, LF-chickens and to human volunteers with low background of lactobacilli,) and determined cell numbers before (day 0) and daily for 5 days after administration by quantitative culture on LRIM plates. As shown in Fig. 2.3, L. reuteri became detectable between

6.0 – 8.0 log10 CFU/g of feces (or per swab) in all hosts within one day of oral administration. In ex-germ-free mice, ex-LF-chickens and ex-germ-free pigs, L. reuteri established populations reaching numbers comparable to those of Lactobacillus in conventional animals (Tannock 1992;

Leser et al. 2002; Abbas Hilmi et al. 2007). In contrast, L. reuteri became detectable in human fecal samples at around 7.0 log10 CFU/g at day 1 and 2 and subsequently followed by a continuous decline, decreasing to numbers that were just above those of the baseline by day 5.

These results were expected as similar temporal patterns were previously observed for the persistence of L. reuteri in humans (Frese, Hutkins and Walter 2012; Rattanaprasert et al. 2014).

Table 2.2 L. reuteri genomes used for phylogeny reconstruction and comparative genomics

Strain (alterative name) Origin Lineage Source mlc3 mouse III JGI: 2506381016 100-23 rat III JGI: 2500069000

LTH2584 sourdough I JGI: 2534682349 TMW1.112 sourdough III JGI: 2534682347

TMW1.656 sourdough III JGI: 2534682350

I5007 pig IV GenBank: CP006011-CP006017 ATCC 53608 pig IV EMBL: LN906634 lp167-67 pig IV JGI: 2599185361

ZLR003 pig N.D JGI: 2687453552 pg-3b pig IV JGI: 2599185334

3c6 pig V JGI: 2599185333

20-2 pig V JGI: 2599185332

TD1 rat I GenBank: CP006603 lpuph1 mouse I JGI: 2506381017 LTH5448 sourdough I JGI: 2571042361

T JCM1112(DSM20016 /F275) human II NCBI: NC_01060 MM4-1a (ATCC PTA-6475) human II JGI: 2502171170

MM2-3 (ATCC PTA-4659) human II JGI: 2502171171

CSF8 chicken VI JGI: 2684623009

1366 chicken VI JGI: 2684623010 JCM1081 chicken VI JGI: 2684623011

CF48-3A human VI JGI: 2502171173 M27U15 human VI JGI: 2687453659

MM34-4A human VI JGI: 2660238834

SD2112 (ATCC 55730) human VI NCBI: NC_015697.1 a as determined in (Oh et al. 2010). N.D not determined

Figure 2.2- Graphic representation of the experimental design. Eighteen strains of different host origin and phylogenetic lineage were grouped into 3 different inocula containing 6 strains each. To facilitate differentiation, strains within the same inoculum were selected to carry distinct leuS alleles. Standardized inocula were prepared to contain equivalent cell numbers of each strain and administered to germ-free mice (n=5 per inocula), Lactobacillus-free chickens (n=5 per inocula), germ-free pigs (n=3 per inocula) and humans with a low background of lactobacilli (n=5 per inocula). Bacteria were cultured from the inoculums and from fecal samples collected between day 1 to 5 after administration, and strain composition was determined by randomly typing colonies

Figure 2.3 Cell numbers of L. reuteri in fecal samples (mice, pigs, humans) or cloacal swabs (chickens) determined by quantitative culture Data are presented as log10 CFU. Each data point represents a sample from individual animals or human volunteers and horizontal bars represent mean ± SD values.

2.3.3 Rodent isolates become enriched in the murine host

Host adaptation of L. reuteri strains to mice had been previously established (Oh et al. 2010;

Frese et al. 2011, 2013). Therefore, L. reuteri strains from rodents were expected to become enriched in this model. As shown in Fig. 2.4, by day 2 (inocula A and B) or day 3 (inoculum C) the relative abundance of rodent strains was significantly higher (p< 0.05) respective to non-rodent lineages. Notably, in those animals receiving inoculum A, nearly all of the colonies typed from days 2 to 5 were identified as the rat isolate CR from the rodent-I lineage. The relative amounts of colonies typed per individual mouse by day 5 are shown in the adjacent bar graphs. As expected, the strain CR was significantly enriched (p< 0.05) in mice that received inoculum A.

Furthermore, 66% of the colonies recovered on day 5 from mice administered inoculum B, belonged to rodent lineages and compared to human-II, poultry-VI and porcine-V strains, the rodent-I strain lpuph1 was significantly enriched (p<0.05). Following the same trend, the rat isolate

N2D of inoculum C represented 67% of the colonies typed on this day, being significantly higher

(p<0.05) than strains from non-rodent lineages.

Together, these results demonstrated that administration of a mixture of strains and subsequent molecular typing (by sequencing the leuS gene) of random colonies from fecal cultures allowed us to determine which L. reuteri strains became enriched under competitive conditions. Thus, this experimental approach can be used to make accurate inferences about the ecological performance of different strains in vivo.

2.3.4 The chicken host

Molecular typing of L. reuteri colonies grown from cloacal swabs of chickens that received inoculum A indicated that from days 3 to 5, the relative abundance of the poultry-VI strain CSF8 was significantly higher (p<0.05) than strains belonging to all other lineages (Fig. 2.4). Similarly, in animals that received inoculum B, the poultry-VI strain 1366 represented between 50 and 70%

of the colonies typed from days 2 to 5. At day 3 the relative abundance of 1366 was significantly higher (p<0.05) than all other strains except the human-VI strain CF483A1. The poultry-VI strain

JCM1081 did not become significantly enriched; however, it represented between 33 and 55% of the colonies recovered from chickens administered inoculum C. By day 5, as shown in the adjacent bar graphs, the strain CSF8 represented a significantly higher (p<0.05) percent of the colonies typed per animal. Similarly, 1366 was found in significantly higher (p<0.05) compared to the human-II and rodent I and III strains. In animals that received inoculum C, the relative abundance of JCM1081 at day 5 (53%) was significantly higher (p<0.05) than most other lineages, except the human-II lineage.

Overall this analysis revealed that chicken-VI strains were enriched during colonization of the chicken gut. Interestingly, we also found that human–VI strains were good colonizers of this host, especially during the early colonization phase. For example, in birds receiving inoculum A, the relative abundance of poultry-VI and human-VI was not significantly different and together these strains accounted for 80% (42.5 % CSF8 and 37.5% M27U15) of the colonies typed on day 1 and

90% (55 % and 36%) on day 2. No significant difference between the relative abundance of poultry-VI strain 1366 and human-VI strain CF484A1 of inoculum B was found at days 1, 3 and

5. Additionally, at day 1, the relative abundance of the human-VI isolate MM344A from inoculum

C was slightly higher (43%) than the poultry isolate JCM1081 (33 %).

Stacked bar plots showing the relative abundance of L. reuteri strains in the inocula and feces of mice, pigs and humans and cloacal swabs of chickens at baseline and during days 1 to 5 after oral administration of 3 different inocula (A, B and C). An asterisk denotes when a strain from host-specific lineage is significantly higher (p<0.05) than all strains from other lineages in the native host. A white star denotes when the percentage of poultry-VI and the human-VI strains is not significantly different (p<0.05) in chickens. A triangle is shown when a strain become 70 significantly enriched in a non- native host. Adjacent bar graphs show the mean and SE of the relative strain abundance of each strain from the colonies typed at day 5. Individual data points represent the percent colonies typed in each animal or human volunteers. Groups labeled with different letters are significantly different (p<0.05). Statistical significance was determined by Figure 2.4 Stacked bar plots showing the relative abundance of L. reuteri one-way ANOVA, α = 0.05.

Figure 2.4 (cont). Stacked bar plots showing the relative abundance of L. reuteri

2.3.5 The pig host

Contrary to results in mice and chickens, strains of pig origin (porcine-VI) did not become significantly enriched (p<0.05) in pigs at any point during the experimental period (Fig. 2.4).

Although some rodent isolates became significantly enriched, this trend was not consistent across inocula, ruling out a competitive advantage of rodent isolates. For example, the strain CR from inoculum A was found in significantly higher amounts (p<0.05) from days 2 to 5 but the strains lpuph1 (inoculum B) and N2D (inoculum C) were only significantly increased (p<0.05) at day 2.

Overall, these results indicate that porcine strains do not possess a competitive advantage in their original host.

2.3.6 The human host

Human-II strains were present in the feces of human volunteers during the 5 day post- inoculation period (Fig 2.4) but were not significantly enriched. The only strains reaching significant enrichment (p<0.05) were of porcine origin, however a competitive advantage of these strains can be ruled out as the trend is not consistent across inocula and only observed for one or two days. For example, the porcine strain jw2015 was enriched only at day 5 in volunteers that consumed inoculum A and the strain ATCC 53508 on days 1 and 2 in volunteers that consumed inoculum C. Notably, human-VI strains were essentially outcompeted in the human host. These findings indicate that L. reuteri strains originating from humans, regardless of their lineage, did not show elevated levels of colonization in the human gut.

2.3.7 Evolution and genome characteristics of lineage-VI strains

To gain a deeper understanding of the evolution of lineage-VI strains, we reconstructed the ancestral states within the L. reuteri phylogeny (Fig. 2.5).This analysis revealed that rodent lineages date back as far as 2 million years, predating all other lineages by at least 800,000 years.

The human lineage-II emerged 1.5 million years ago from a rodent ancestor (Event 1). A host

switch approximately 1 million years ago resulted in the emergence of the poultry lineage-VI and porcine lineage-IV (Event 2). The porcine lineage-V appeared much later 96,000 years ago (Event

3). Most importantly, all isolates of lineage-VI share a common ancestor associated with poultry, with the human isolates emerging latest, less than 61,000 years ago (Event 4) (Fig. 2.5).

Next we determined the genomic characteristics of lineage-VI strains by analyzing the genomes of three strains originated from chicken (1366, CSF8 and JCM108) and four strains of human origin (SD2112, MM34-4A, M27U15 and CF483A). As shown in Fig. 2.6 all seven analyzed lineage-VI strains shared 1433 predicted orthologous genes. Beyond this core genome, poultry-VI strains possessed an open pangenome with large number of strain specific genes (197 genes in 1366, 215 in CSF8 and 484 in JCM1081) and an average nucleotide identity (ANI) between 98.77% and 99.06%. This is in stark contrast to human-VI strains which showed very few unique genes and essentially a closed pangenome (between 3 - 6 genes per strain) and an

ANI between 99.92% and 99.99%. These findings are in line with the tight clustering of human-IV strains in the phylogenetic tree shown in Fig. 2.1.

We then sought to identify genes specific to (not present in other L. reuteri genomes) and conserved in the genomes of all seven strains in the lineage-VI. We found only 3 genes which encoded an aspartate racemase (EC 5.1.1.13) and two transcriptional regulators of the XRE and

DeoR families. Next we identified 28 genes that were conserved in the poultry-VI and absent in the genomes of human-VI strains. Of those 10 were annotated as hypothetical proteins, 4 as phage-related, three as c-di-GMP activations via the GGDEF-EAL transduction system involved in biofilm formation (Gjermansen, Ragas and Tolker-Nielsen 2006), and one as a transcriptional regulator of the HxlR family. The remaining encoded proteins involved in vitamin biosynthesis or sugar transporters/transferases. Notably, none of these genes appears to be exclusive to poultry-

VI strains when compared to other L. reuteri genomes (Table 2.3)

Figure 2.5 Ancestral state analysis Inferred evolutionary history of L. reuteri-host associations. Ancestral states were inferred on bacterial phylogeny modified from a previous study (Oh et al. 2010). The tree is a maximum likelihood reconstruction of a concatenated set of 7 single-copy genes from 116 strains. Colors represent host state on the tips of the tree and inferred states on ancestral nodes. Equivocal ancestral states are represented by mixed colors in the circle. Four host switching events were highlighted as enlarged circles (labeled 1–4). The time scale is in scale of thousand years (kyrs) estimated using Bayesian phylogenetic analyses.

Figure 2.6 Core and strain-specific gene content of L. reuteri lineage-Vi strains

Each oval represents the genomes of poultry-VI (yellow) and human-VI (orange) strains. The core gene set is indicated by the number in the center and below in parenthesis is number of unique lineage-VI genes when compared with other L. reuteri strains. The outer values in the ovals indicate the number of genes unique to each strain and shown below in parentheses are the numbers of strain-specific genes after subtracting genes found in other L. reuteri genomes

Next we examined whether human-VI strains possessed genes not present in poultry-VI strains. Of the 190 proteins identified through this process, 92 were annotated as hypothetical proteins, and 61 as transposases, endonucleases or phage related. Genes with functional annotation that did not fall in these categories are listed in Table 3. Interestingly, 13 of the genes are co-localized in a putatively horizontally-acquired (flanked by transposase genes and 10% higher GC content than the genome average) gene cluster that encodes membrane glycosyl transferases and transmembrane transporters predicted to be involved in capsular polysaccharide biosynthesis. Another interesting human-VI-specific gene included a tetracycline resistance cassette, encoded in plasmid pLR581 previously described for strain L. reuteri SD2112

(Rosander, Connolly and Roos 2008). Human-VI strains also harbor three additional plasmids encoding lincomycin resistance (pLR585), cadmium resistance (pLR584), and uncharacterized functions (pLR580). Not all the genes encoded in these plasmids are exclusive to human-VI

strains. However, the replication proteins are conserved in all of the human-VI genomes surveyed

(Table 2.3) and the presence of these plasmids was confirmed via extraction and agarose gel visualization (data not shown).

Table 2.3 Genes-specific to Poultry-VI and Human-VI strains Specific in poultry-VI Specific in human-VI

IDa Gene name or function IDa Gene name or function 20805 GGDEF c-di-GMP synthetase 64061 dTDP-glucose 4,6-dehydratase /epimerase 20806 EAL c-di-GMP phosphodiesterase 64063 CDP-glycerol phosphotransferase 20821 Cytochrome Heme/Steroid domain 64064 Glycosyltransferase 21815 Butanediol dehydrogenase 64065 Wzx-flippase 21816 Sugar phosphate permease 64066 Transmembrane protein 21916 UDP-glycosyltransferase 64067 Glycosyltransferase 22288 Xanthine/uracil /vitC permease 64068 Glycosyltransferase 22433 Thiamine phosphate synthase 64069 Glycosyltransferase 22434 Hydroxymethylpyrimidine kinase 64070 1,6-galactosyltransferase 22959 HxlR transcriptional regulator 64071 Glycosyltransferase 22963 Sugar transferase LPS biosynthesis 64072 Glycosyltransferase 23007 Hydroxyethylthiazole kinase 64073 Glycosyltransferase

64074 Nucleoside-diphosphate-sugar epimerase

64355 O-acyltransferase

64358 Glycosyltransferase

64361 UDP-galactofuranosyltransferase

64432 SNF2 family helicase

64483 ABC transporter ATP-binding protein

64484 ABC transporter

64485 LytTR transcriptional regulator

64759 Glucansucrase

65314 PglZ alkaline-phosphatase

65474 Lantibiotic protection ABC transporter

Table 2.3 (cont)

65587 MarR transcriptional regulator

66329 CAAX amino protease (plasmid pLR585)

66330 MerR transcriptional regulator (pLR585)

66332 Replication protein (pLR585)

66333 Replication associated protein (pLR585)

66334 Replication associated protein (pLR585)

66336 Replication protein (pLR580)

66348 Tetracycline resistance protein Tet (pLR581)

66357 Initiator RepB protein (pLR581)

66364 Replication rep protein (pLR584)

66370 Lysophospholipase family protein (pLR584)

66374 Transcriptional repressor CopY (pLR584) aGene ID as assigned by JGI to L. reuteri SD2112 and L. reuteri JCM1081 for the genes specific to poultry-VI human-VI, respectively. Genes IDs in bold are exclusive to human-VI strains and genes in are italics are co-localized at the Wzx-dependent capsular polysaccharide synthesis cluster. For non- chromosomal genes the name of the harboring plasmids is shown in parenthesis

2.4 Discussion

In this study, we devised an experimental approach that allowed us to directly assess the competitive interactions among L. reuteri strains in the gastrointestinal tracts of different vertebrate hosts. By testing for the relative enrichment of a specific strain, the assay directly compares the ability of strains to propagate under the ecological conditions of the gut. Since population growth directly relates to ecological fitness of microbes (Garcia and Gerardo 2014), an enrichment of strains from a particular phylogenetic lineage, in our models, does provide direct evidence for adaptedness. An interpretation of the findings can therefore be used to infer the evolutionary history of a phylogenetic lineage and the underlying mechanism that drove its diversification.

We tested this approach and confirmed host adaption of rodent isolates to the mouse gut. As shown previously (Oh et al. 2010; Frese et al. 2011), colonization phenotypes were in agreement with phylogeny as both mouse and rat strains show similar ecological performance in mice. This suggests that rodent lineages are not evolutionarily confined to a specific host species but are adapted to specific gastrointestinal conditions that are expected to be very similar in mice, rats and possibly other rodents (Oh et al. 2010). Similar trends have been observed for the bumble bee symbiont Snodgrassella alvi which appears to have adapted to the host genera (Bombus) and not the species (Kwong and Moran 2015). Although much remains unknown regarding the dynamic evolutionary patterns and adaptive paths of L. reuteri strains within rodents, our findings in mice indicate that competition experiments with standardized mixtures of strains are appropriate to make accurate assessments of host adaption, and therefore allow insight on the evolutionary outcomes of host-microbe interrelationships.

2.4.1 Lineage VI strains, even if isolated from humans, show elevated fitness in chicken

In agreement with previous studies (Oh et al. 2010; Spinler et al. 2014), the genome-wide phylogenetic analysis of L. reuteri showed that strains isolated from poultry cluster together with a subset of human isolates into one cohesive phylogenetic lineage (lineage-VI). Although we have previously speculated about the natural history of this lineage (Oh et al. 2010), interpretation of phylogenies alone can lead to incorrect conclusions (Moran and Sloan 2015). Our experiments here, now demonstrate that poultry-VI strains are host adapted to chickens and that human-VI strains also performed well in this host, while showing an extremely low performance in humans.

Accordingly, the genomic analysis revealed that human-VI isolates are likely to possess all the colonization factors necessary to colonize the chicken gut, as only very few genes were present in poultry-VI strains were absent from human-VI strains (Table 3). Both poultry and human lineage-VI strains possess a lineage-specific putative regulatory system, suggesting an important role of environmental sensing in the colonization of the chicken gastrointestinal tract. In this

regard, a single regulatory gene has been shown to alter host-specificity in Vibrio fischeri (Mandel et al. 2009) and rodent-specific two component systems are known to regulate biofilm formation of L. reuteri in mice (Frese et al. 2011, 2013). These findings indicate that human-VI isolates, alike poultry-VI isolates, share an evolutionary history with poultry. Accordingly, the immediate ancestral node of human-VI strains was inferred to be associated with poultry (Fig 5).

The above-mentioned findings beg the question of why lineage-VI strains have been isolated from humans if the lineage has maintained stringent evolutionary ties with poultry and are host- adapted? One possibility is that L. reuteri switched from poultry to humans and became permanently associated with this new host. However, in our human experiment, human-VI strains were completely outcompeted, thus ruling out an adaptation to the human gastrointestinal tract.

In addition, the majority of the human-VI isolates originate from extra-intestinal sources (breast milk, vagina, mouth) for which the species L. reuteri has not been described as a significant member (Ravel et al. 2011; Jost et al. 2013), which is contrary to poultry where L. reuteri is autochthonous and an abundant member of the gut microbiota (Abbas Hilmi et al. 2007). Based on these considerations, we propose that specific strains from lineage VI can become transiently associated with humans. Microbial exchange between avian and humans is not only possible but frequent, as demonstrated by 2.5 million cases per year of food borne illnesses in the United

States that arise from the transmission of pathogens from poultry, meat, and eggs in the United

States (Painter et al. 2013). Interestingly, our genomic analysis revealed that poultry-VI strains possess a large and adaptable pangenome, while human-VI strains show very little genomic variation. These findings suggest that essentially one single clone (< 375 SNPs when compared to L. reuteri SD2112) amongst the lineage-VI has been repeatedly isolated from humans. Similar phenomena have been observed for globally spread monomorphic pathogens (Achtman 2012), such as Yersinia pestis in which the acquisition of a few horizontally acquired traits was sufficient to enable switching hosts (Achtman et al. 1999). Under this scenario, and given the high number

of human-VI specific mobile genetic elements found in the genomes of human-VI strains, it is possible that this clone acquired specific traits that allowed a temporary migration to humans. The

Wzx-dependent capsular polysaccharide biosynthetic gene cluster and the plasmid-encoded antibiotic resistance cassette, which are both likely to be horizontally acquired, might represent key traits that allow transfer to humans. The ability of these strains to grow in the presence of tetracycline was confirmed experimentally (data not shown) and capsular molecules can induce or suppress host immune responses (Comstock and Kasper 2006; Willis and Whitfield 2013).

Under this perspective, it is tempting to speculate about a scenario in which a specific L. reuteri clone was able evade the immune system and temporality colonize the human host, presumably after a course of antibiotics.

2.4.2 Human and porcine isolates do not show elevated ecological fitness in their respective hosts

Human-II strains cluster separately from all other L. reuteri strains. This lineage is both remarkably homogeneous and specific to humans (Oh et al. 2010; Walter, Britton and Roos

2011). However, our experiment did not provide sufficient evidence of adaptation of these strains to humans, even though they persisted more than human-VI strains. This finding could suggest that the evolution of the human-II lineage was not driven by specialized adaptations to the human gut. Another possible explanation is that these isolates have a non-human niche (i.e., environmental, food, other hosts). However, the phylogenetic cohesiveness of these strains argues strongly against this scenario. It is widely documented that food and plant-derived lactobacilli strains are commonly isolated from human fecal samples but these strains, unlike L. reuteri human-II strains, are neither phenotypically nor phylogenetically related (Siezen et al.

2010; Smokvina et al. 2013; Martino et al. 2016). We cannot rule out that the human resident microbiota could have prevented the experimental L. reuteri strains from becoming established.

Although recent work provided evidence that a gut microbe can engraft in the human gut if an

open niche is available (Maldonado-Gómez et al. 2016). In this respect, it is important to point out that L. reuteri was commonly detected as a member of the human gut microbiota according to studies conducted between 1950 and 1960 (Reuter 2001), while it is rarely detected among the human gut microbiota today (Walter 2008; Qin et al. 2010; Martínez, Muller and Walter 2013).

This suggests that L. reuteri might have been displaced as a dominant member of the human gut microbiota due to environmental changes (e.g., antibiotic usage, hygiene and dietary practices) associated with modern lifestyles (Walter, Britton and Roos 2011) as it has been proposed for other members of the microbiota (Blaser and Falkow 2009). Although speculative, this idea is supported by a recent study in which L. reuteri was found to be a dominant member of the microbiota of rural Papua New Guineans (Martínez et al. 2015). Under this perspective, the absence of the niche in which human-II L. reuteri evolved, would have prevented these strains from becoming enriched in our human experiment. Further work will be required to resolve outstanding questions regarding the evolution of L. reuteri with humans.

Contrary to the situation in modern humans, L. reuteri is a core member and one of the most abundant Lactobacillus species of the pig’s microbiota (Leser et al. 2002), suggesting L. reuteri is symbiont of pigs. However, a recent genome-wide comparative study did not find any host- specific genomic signatures among porcine isolates that cluster into two pig-confined phylogenetic clades (Oh et al. 2010; Wegmann et al. 2015). This finding suggests that clustering of these strains could be driven by neutral evolution or ecological factors not directly related to the porcine host. Our data agrees with this notion as we found no evidence for host adaptation in pigs. Nevertheless, it is important to point out that our pig model is not exempt of limitations that could explain these observations. For example it is conceivable that complex interactions with the microbiota (both cooperative and antagonistic) might underlie the adaptation of L. reuteri to pigs.

L. reuteri is one member of a multi-species biofilm that forms on the squamous keratinized lining of the pig’s pars esophageal tissue (61). In this sense, the germ-free pig model might fail to

recapitulate adaptive interactions between L. reuteri and co-existing members of the biofilms, such as L. amylovorus and L. johnsonii, which are the dominant Lactobacillus species in pigs

(Leser et al. 2002). Overall, our findings suggest a direct co-evolution of L. reuteri with rodents and chicken (potentially as a primary colonizer in the biofilms), while in pigs, the species may evolve in a tripartite inter-relationship with the host and other microbes.

Another important factor to consider is that dietary glycans can have a direct effect in the composition of the pig’s microbiota, including the abundance of lactobacilli (Frese et al. 2015).

The infant formula fed to piglets contained galactooligosaccharides (GOS) that are utilized by almost all L. reuteri strains independent of host origin (Goin 2010). It is possible that porcine L. reuteri strains have evolved to utilize oligosaccharides in the milk of pigs and that this adaptation resulted in the host-confined phylogenetic clusters. A formula in which these carbohydrates are replaced by the non-selective GOS would have rendered strain-specific differences in the ability to utilize milk carbohydrates insignificant, potentially removing the ecological advantage of porcine isolates. Additional animal experiments will ultimately be necessary to derive clear conclusions on the existence of host adaptation of L. reuteri to pigs and the mechanisms by which the porcine- specific lineages arose.

2.4.3 Limitations of this study

It is extremely difficult to experimentally replicate the ecological conditions under which bacteria evolve, which are often dynamic and subject to change. The animal host is an excellent replication of the natural habitat of a gut symbiont, but as described above, both germ-free and conventional models have limitations. Germ-free models fail to replicate the interrelationships between members of the community, which might be especially relevant in the evolution of L. reuteri as a component of biofilms. On the other hand, in hosts with a conventional microbiota, such as the humans in our study, the niche might already be occupied by more resilient taxa, thus

preventing the establishment of external strains. Future studies should be devoted to apply this experimental system to more refined models such as gnotobiotic animals, especially containing species that co-exist with L. reuteri in natural settings.

2.5 Conclusion

Results from this study contribute to our understanding about the evolutionary history of the gut symbiont L. reuteri, which is now the first vertebrate gut symbiont for which specific adaptation has been experimentally proven in different hosts. This finding suggests that long periods of strict host-association is required for functional adaptation. This work also expands our knowledge about the various lifestyles and the array of selective pressures shaping the evolution of

Lactobacillus species. For example, the adapted lifestyle of L. reuteri sharply contrasts with that of the generalists L. plantarum (Martino et al. 2016), L. paracasei (Smokvina et al. 2013) and L. rhamnosus (Douillard et al. 2013). This aspect can be particularly important in the selection of probiotics, as functional attributes can be directly related to the evolution of particular strains and the nature of the symbiotic relationship maintained with different hosts (Spinler et al. 2014).

Survival and persistence in the digestive system might also be a desirable trait for some probiotic applications. Therefore, our findings also provide an ecological and evolutionary basis for the selection of strains for probiotic applications in different hosts.

2.6 Acknowledgments

We would like to thank all the volunteers who participated in this study. We are also grateful to Dr. Huyen Tran and Vicky Samek for assistance with the pig experiment and to Janis Cole for assistance with the human trial. J.B acknowledges the BMBF grant FKZ 031A533 from the de.NBI network. C.M.S is an Alberta Innovates Technology Futures iCORE Strategic Chair in Bacterial

Glycomics and J.W acknowledges support by BioGaia AB and the Campus Alberta Innovates

Program.

2.6.1 Funding information

This study was funded in part by the Natural Sciences and Engineering Research Council of

Canada (NSERC) Discovery Grant awarded to J.W.

2.7 References

Abbas Hilmi HT, Surakka A, Apajalahti J et al. Identification of the Most Abundant

Lactobacillus Species in the Crop of 1- and 5-Week-Old Broiler Chickens. Appl Environ Microbiol

2007;73:7867–73.

Achtman M. Insights from genomic comparisons of genetically monomorphic bacterial pathogens. Philos Trans R Soc B Biol Sci 2012;367:860–7.

Achtman M, Zurth K, Morelli G et al. Yersinia pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis. Proc Natl Acad Sci U S A 1999;96:14043–8.

Bankevich A, Nurk S, Antipov D et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 2012;19:455–77.

Blaser MJ, Falkow S. What are the consequences of the disappearing human microbiota? Nat

Rev Microbiol 2009;7:887–94.

Blom J, Kreis J, Spänig S et al. EDGAR 2.0: an enhanced software platform for comparative gene content analyses. Nucleic Acids Res 2016;44:W22–8.

Bongrand C, Koch EJ, Moriano-Gutierrez S et al. A genomic comparison of 13 symbiotic Vibrio fischeri isolates from the perspective of their host source and colonization behavior. ISME J

2016;10:2907–17.

Bouckaert R, Heled J, Kühnert D et al. BEAST 2: A software platform for bayesian evolutionary analysis. PLoS Comput Biol 2014;10, DOI: 10.1371/journal.pcbi.1003537.

Comstock LE, Kasper DL. Bacterial glycans: key mediators of diverse host immune responses. Cell 2006;126:847–50.

Dale C, Moran NA. Molecular interactions between bacterial symbionts and their hosts. Cell

2006;126:453–65.

Darling AE, Mau B, Perna NT. progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement. PLoS One 2010;5:e11147.

Dethlefsen L, McFall-Ngai M, Relman DA. An ecological and evolutionary perspective on human-microbe mutualism and disease. Nature 2007;449:811–8.

Doolittle WF, Zhaxybayeva O. On the origin of prokaryotic species. Genome Res

2009;19:744–56.

Douillard FP, Ribbera A, Kant R et al. Comparative Genomic and Functional Analysis of 100

Lactobacillus rhamnosus Strains and Their Comparison with Strain GG. Richardson PM. PLoS

Genet 2013;9:e1003683.

Falush D, Wirth T, Linz B et al. Traces of Human Migrations in Helicobacter pylori populations.

Science 2003;299:1582-85

Fraune S, Bosch TCG. Why bacteria matter in animal development and evolution. BioEssays

2010;32:571–80.

Frese SA, Benson AK, Tannock GW et al. The evolution of host specialization in the vertebrate gut symbiont Lactobacillus reuteri. Guttman DS (ed.). PLoS Genet 2011;7:16.

Frese SA, Hutkins RW, Walter J. Comparison of the colonization ability of autochthonous and allochthonous strains of lactobacilli in the human gastrointestinal Tract. Adv Microbiol

2012;2:399–409.

Frese SA, MacKenzie DA, Peterson DA et al. Molecular Characterization of Host-Specific

Biofilm Formation in a Vertebrate Gut Symbiont. Garsin DA (ed.). PLoS Genet 2013;9:e1004057.

Frese SA, Parker K, Calvert CC et al. Diet shapes the gut microbiome of pigs during nursing and weaning. Microbiome 2015;3:28.

Garcia JR, Gerardo NM. The symbiont side of symbiosis: do microbes really benefit? Front

Microbiol 2014;5:510.

Gjermansen M, Ragas P, Tolker-Nielsen T. Proteins with GGDEF and EAL domains regulate

Pseudomonas putida biofilm formation and dispersal. FEMS Microbiol Lett 2006;265.

Goin CD. Growth of lactic acid bacteria: infuence of protocooperation, bacteriophage infection, and prebiotic carbohydrates. 2010.

Hall T. BioEdit: a user-friendly biological sequence alignment editor and analysis program for

Windows 95/98/NT. Nucleic Acids Symp Ser 1999;41:95–8.

Jost T, Lacroix C, Braegger C et al. Assessment of bacterial diversity in breast milk using culture-dependent and culture-independent approaches. Br J Nutr 2013;110:1253–62.

Kwong WK, Engel P, Koch H et al. Genomics and host specialization of honey bee and bumble bee gut symbionts. Proc Natl Acad Sci 2014;111:11509–14.

Kwong WK, Moran NA. Evolution of host specialization in gut microbes: the bee gut as a model. Gut Microbes 2015;6:214–20.

Lawley TD, Walker AW. Intestinal colonization resistance. Immunology 2013;138:1–11.

Leser TD, Amenuvor JZ, Jensen TK et al. Culture-Independent Analysis of Gut Bacteria: the

Pig Gastrointestinal Tract Microbiota Revisited. Appl Environ Microbiol 2002;68:673–90.

Ley RE, Hamady M, Lozupone C et al. Evolution of mammals and their gut microbes. Science

2008a;320:1647–51.

Ley RE, Lozupone C a, Hamady M et al. Worlds within worlds: evolution of the vertebrate gut microbiota. Nat Rev Microbiol 2008b;6:776–88.

Maddison W, Maddisson D. Mesquite: A modular system for evolutionary analysis. Version

3.2. 2017.

Maldonado-Gómez MX, Martínez I, Bottacini F et al. Stable engraftment of Bifidobacterium longum AH1206 in the human gut depends on individualized features of the resident microbiome.

Cell Host Microbe 2016;20:515–26.

Mandel MJ, Wollenberg MS, Stabb E V et al. A single regulatory gene is sufficient to alter bacterial host range. Nature 2009;458:215–8.

Markowitz VM, Szeto E, Palaniappan K et al. The integrated microbial genomes (IMG) system in 2007: data content and analysis tool extensions. Nucleic Acids Res 2008;36:D528-33.

Martínez I, Muller CE, Walter J. Long-term temporal analysis of the human fecal microbiota revealed a stable core of dominant bacterial species. Gilbert JA. PLoS One 2013;8:e69621.

Martínez I, Stegen JC, Maldonado-Gómez MX et al. The Gut Microbiota of Rural Papua New

Guineans: Composition, Diversity Patterns, and Ecological Processes. Cell Rep 2015;11:527–38.

Martino ME, Bayjanov JR, Caffrey BE et al. Nomadic lifestyle of Lactobacillus plantarum revealed by comparative genomics of 54 strains isolated from different habitats. Environ Microbiol

2016;18:4974–89.

Miniats OP, Jol D. Gnotobiotic pigs-derivation and rearing. Can J Comp Med Rev Can mé decine Comp 1978;42:428–37.

Moeller AH, Caro-Quintero A, Mjungu D et al. Cospeciation of gut microbiota with hominids.

Science 2016;353:380-382.

Moran NA. Symbiosis. Curr Biol 2006;16:R866–71.

Moran NA, Sloan DB. The Hologenome Concept: Helpful or Hollow? PLOS Biol

2015;13:e1002311.

Ochman H, Worobey M, Kuo C-H et al. Evolutionary relationships of wild hominids recapitulated by gut microbial communities. Achtman M (ed.). PLoS Biol 2010;8:e1000546.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Painter JA, Hoekstra RM, Ayers T et al. Attribution of foodborne illnesses, hospitalizations, and deaths to food commodities by using outbreak data, United States, 1998-2008. Emerg Infect

Dis 2013;19:407–15.

Pot B, Felis GE, Bruyne K De et al. The genus Lactobacillus. Lactic Acid Bacteria. Chichester,

UK: John Wiley & Sons, Ltd, 2014, 249–353.

Qin J, Li R, Raes J et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010;464:59–65.

Rattanaprasert M, Roos S, Hutkins RW et al. Quantitative evaluation of synbiotic strategies to improve persistence and metabolic activity of Lactobacillus reuteri DSM 17938 in the human gastrointestinal tract. J Funct Foods 2014;10:85–94.

Ravel J, Gajer P, Abdo Z et al. Vaginal microbiome of reproductive-age women. Proc Natl

Acad Sci U S A 2011:4680–7.

Reuter G. The Lactobacillus and Bifidobacterium microflora of the human intestine: composition and succession. Curr Issues Intest Microbiol 2001;2:43–53.

Rogosa M, Mitchell JA, Wiseman RF. A selective medium for the isolation and enumeration of oral and fecal lactobacilli. J Bacteriol 1951;62:132–3.

Rosander A, Connolly E, Roos S. Removal of antibiotic resistance gene-carrying plasmids from Lactobacillus reuteri ATCC 55730 and characterization of the resulting daughter strain, L. reuteri DSM 17938. Appl Environ Microbiol 2008;74:6032–40.

Rosenberg E, Zilber-Rosenberg I. Microbes Drive Evolution of Animals and Plants: the

Hologenome Concept. MBio 2016;7:e01395.

Sachs J, Skophammer R, Regus J. Evolutioanry transitions in bacterial symbiosis. PNAS

2011;108:10800–7.

Siezen RJ, Tzeneva VA, Castioni A et al. Phenotypic and genomic diversity of Lactobacillus plantarum strains isolated from various environmental niches. Environ Microbiol 2010;12:758–73.

Smokvina T, Wels M, Polka J et al. Lactobacillus paracasei Comparative Genomics: Towards

Species Pan-Genome Definition and Exploitation of Diversity. Highlander SK. PLoS One

2013;8:e68731.

Spinler JK, Sontakke A, Hollister EB et al. From prediction to function using evolutionary genomics: human-specific ecotypes of Lactobacillus reuteri have diverse probiotic functions.

Genome Biol Evol 2014;6:1772–89.

Stevens CE, Huma ID. Contributions of microbes in vertebrate gastrointestinal tract to production and conservation of nutrients. Physiol Rev 1998;78:393–427.

Tannock GW. The lactic microflora of pigs, mice and rats. The Lactic Acid Bacteria Volume 1.

Boston, MA, MA: Springer US, 1992, 21–48.

Walter J. Ecological role of lactobacilli in the gastrointestinal tract: implications for fundamental and biomedical research. Appl Environ Microbiol 2008;74:4985–96.

Walter J, Britton RA, Roos S. Host-microbial symbiosis in the vertebrate gastrointestinal tract and the Lactobacillus reuteri paradigm. Proc Natl Acad Sci U S A 2011;108 Suppl:4645–52.

Wattam AR, Abraham D, Dalay O et al. PATRIC, the bacterial bioinformatics database and analysis resource. Nucleic Acids Res 2014;42:D581-91.

Wegmann U, MacKenzie DA, Zheng J et al. The pan-genome of Lactobacillus reuteri strains originating from the pig gastrointestinal tract. BMC Genomics 2015;16:1023.

Willis LM, Whitfield C. Structure, biosynthesis, and function of bacterial capsular polysaccharides synthesized by ABC transporter-dependent pathways. Carbohydr Res

2013;378:35–44.

Wollenberg MS, Ruby EG. Phylogeny and fitness of Vibrio fischeri from the light organs of

Euprymna scolopes in two Oahu, Hawaii populations. ISME J 2012;6:352–62.

3. Chapter three: A rodent-strain specific two- component system is involved in biofilm formation of

Lactobacillus reuteri 100-23.

3.1 Introduction

The ability of bacteria to evolve and adapt to new habitats is often associated with changes in gene content, genetic regulation and consequently in their functional capacity (Altermann, 2012).

These effects are imposed by constraints encountered in the ecosystems they inhabit and adapt to (Alm, Huang and Arkin 2006; Toft and Andersson 2010). The bacterial species Lactobacillus reuteri inhabits the gastrointestinal tract (GI) of vertebrate animals (Tannock 1992; Leser et al.

2002; Abbas Hilmi et al. 2007) and shares an evolutionary history with pigs, rodents, poultry and humans, that resulted in the emergence of phylogenetic lineages associated with host origin (Oh et al. 2010; Walter, Britton and Roos 2011; Duar et al. 2017 (Chapter 2)). The evolution of L. reuteri with rodents was adaptive and resulted in host specificity (Oh et al. 2010; Frese et al. 2013;

Duar et al. 2017 (Chapter 2)). L. reuteri are dominant members of the murine microbiota and maintain high populations throughout the life of the animal, in part facilitated by their ability to form biofilms in the mouse forestomach, a trait that is exclusive to rodent isolates (Frese et al.

2013). The genomes of rodent strains encode a series of functions that reflect adaptations to the conditions in the mouse forestomach and are likely to be determinants of host-specificity. These factors include large adhesin-like surface proteins, SecA2-SecY2 accessory secretion systems, a urease cluster and two-component regulatory systems (Frese et al. 2011). Most of these genes are exclusively found in the genomes of rodent isolates and are rarely found or absent in strains from other hosts (Frese et al. 2011). Studies with loss-of-function mutants have demonstrated that these genes are elemental to the ecological success of L. reuteri in the mouse GI tract (Frese et al. 2011, 2013; Krumbeck et al. 2016). The urease cluster mediates resistance to acid conditions encountered in the forestomach (Wilson et al. 2014; Krumbeck et al. 2016). Selective epithelial adhesion and subsequent biofilm formation, are initiated through a rodent-specific surface adhesin serine-rich protein with a devoted SecA2-SecY2 transport system (Frese et al.

2013).

Biofilms are sessile layers of bacteria encased in an extracellular matrix composed of exopolysaccharides, proteins and nucleic acids (Branda et al. 2005). Biofilm formation is a highly coordinated mechanism often driven by a response to changing environmental conditions (Abee et al. 2011), and genes involved in the process are often transcriptionally regulated (Davey and

O’toole 2000). In gram positive bacteria, transcriptional regulation of biofilm formation is commonly mediated by two-component regulatory systems (TCS) which typically consist of a membrane-bound histidine kinase (HK) sensor and a cytoplasmic response regulator (RR)

(Monedero, Revilla-Guarinos and Zúñiga 2017). HKs respond to extracellular cues by transferring a phosphate group from a conserved histidine residue to an also conserved aspartate residue in a cognate RR. Phosphorylation of the RR subsequently activates the transcriptional output of the pathway (Capra and Laub 2012). TCSs are involved in the regulation and development of biofilm formation of a wide number of bacterial species and a variety of habitats and conditions (Li et al.

2002; Zhang et al. 2004; Mikkelsen, Sivaneson and Filloux 2011; Su and Gänzle 2014;

Norsworthy and Visick 2015) but more importantly, a specific two-component sensor kinase can determine host-specificity, as it has been demonstrated for the marine symbiont Vibrio fisheri

(Mandel et al. 2009).

In a previous comparative genomics study two different TCS were identified to be highly specific to rodent strains (Frese et al. 2011). One of these systems was characterized by Su and

Gänzle (2014) and found to be relevant to biofilm formation in a substrate dependent fashion (Su and Gänzle 2014). The other TCS (TCS_70529-30) encodes an AgrA-family sensor histidine kinase (HK_70529) and a LytTR transcriptional regulator (RR_70530). Inactivation of an ABC- type transporter located downstream of TCS70529-30 impairs biofilm formation in vivo (Frese et al. 2013). However, the specific role of TCS70529-30 in biofilm formation of L. reuteri remains to be determined. It was therefore the aim of this study to characterize the role of the rodent-specific

TCS70529-30 in L. reuteri using the rat isolate 100-23 as a model. Single gene knockouts of the individual genes were generated, and the ability of deletion mutants to form biofilms in vivo was

determined in ex-germfree mice and compared to the wildtype (WT). To gain further insight into the specific roles of TCS70529-30 in the structural integrity of the biofilm, macroscopic and microstructural differences in the architecture of the biofilm matrix as well as the analysis of the mechanical strength properties, were conducted using in vitro techniques.

3.2 Materials and Methods

3.2.1 Bacterial strains, plasmids and media

The bacterial strains and plasmids used in this study are listed in Table 3.1. L reuteri strains used for this study were derived from the wild-type strain 100-23. L. reuteri strains were grown in modified MRS medium (mMRS) supplemented with 10 g/l maltose and 5 g/l fructose at

37 °C, unless otherwise noted. When antibiotic selection was required, 10 mg/l of ampicillin or 5 mg/l of erythromycin were added to the mMRS media. Escherichia coli DH5α was used for cloning procedures. Ampicillin (100 mg/l) or erythromycin (500 mg/l) was added to Luria-Bertani media as needed for selection of clones.

Table 3.1 Bacterial strains and plasmids used in this study Strain or Plasmid Genotype Reference

Strains Escherichia coli DH5α Cloning host for pUC19 and (Sambrook and W Russell PJRS233-derivative plasmids 2001) Lactobacillus reuteri 100-23 Wildtype (Wesney and Tannock 1979) Δhk70529 Deletion of the hk70529 This study Δrr70530 Deletion of the rr70530 This study

Plasmids pUC19 Cloning vector used in E.coli; (Yanisch-Perron, Vieira and Ampr Messing 1985) pJRS233 Shuttle vector used in the hosts (Perez-Casal et al. 1993) E. coli and L. reuteri 100-23; Ermr Ermr, erythromycin resistance gene ; Ampr , Ampicillin resistance gene

3.2.2 DNA manipulations

Routine molecular biology techniques were performed according to standard procedures

(Sambrook and W Russell 2001). Restriction and modifying enzymes (Thermo Scientific

FastDigest) were used as recommended by the manufacturer. Plasmid DNA was prepared from

E. coli using QIAGEN Miniprep kits. Chromosomal DNA and plasmid DNA were isolated from L. reuteri as previously described in Oh et al. (2010) and Walter et al. (2003), respectively.

3.2.3 Generation of L. reuteri 100-23 knockout mutants

In-frame knockout L. reuteri mutants were generated according to a deletion strategy described previously (Su, Schlicht and Gänzle 2011). Plasmids and primers are listed in Table 2.

Briefly, ~1kb upstream and downstream regions of the target genes were amplified by PCR from genomic DNA, and referred to as flanking regions FR-A and FR-B, respectively. Amplicons of

FR-A and FR-B were inserted separately into pUC19 vector to generate pGeneFR-A and pGeneFR-B. Flanking regions were then digested with the corresponding restriction enzymes and ligated into pUC19 using T4 DNA ligase to produce pGene_FR-AB. The ligated fragments were digested from pGene_FR1-4 and inserted into the shuttle vector PJRS233 (Perez-Casal et al.

1993) to generate the knockout plasmid p-Gene_KO. The resulting knockout plasmid was purified and used to electrotransform (12. kV/cm, 25 µF capacitance, 400 Ώ resistance) 100 µl of competent L. reuteri 100-23 cells. Electrotransformed lactobacilli were incubated in 1.0 ml of pre- warmed mMRS at 37°C for 2.5h to allow phenotypic expression and then plated on mMRS agar plates containing erythromycin (10 µg/ml). After incubation for 24 hours, 10 individual colonies carrying the plasmid were transferred into mMRS-erythromycin broth (10 µg/ml) and incubated at

45°C for 80 generations. Following, single-crossover L. reuteri mutants were cured by culturing in mMRS broth at 37°C for 100 generations. Antibiotic sensitive double-crossover mutants were identified by replica plating onto mMRS and mMRS- erythromycin agar plates. In-frame deletions

were confirmed by PCR and Sanger sequencing using the primer set gene-KO-F and gene-KO-

R (Table 3.2).

Table 3.2 Primers used in this study to generate knockout mutants Primer name Sequence 5’ to 3’ Features

70529HK_KO1_BamHI ATTGGATCCGCAAACAGTAAACGCCAAAA Forward primer for 5’ flaking sequence of hk70529

70529HK_KO2_SalI CCGAATGTCGACTTGTGCTAACGTTAATTGAATCATC Reverse primer for 5’ flaking sequence of hk70529

70529HK_KO3_SalI CCGAATGTCGACACTTGTTTTATTCAGGGAAAGTGAG Forward primer for 3’ flaking sequence of hk70529

70529HK_KO4_HindIII ATGAAGCTTCAAAATTCGTAAGCCTTTCTGC Reverse primer for 3’ flaking sequence of hk70529

70530RR_KO-F TGGTGGATTTTGATTTAGAAACG Forward sequencing primer of Δhk70529

70530RR _KO-R GTGAGTATCCCCATCCTCCA Reverse sequencing primer of Δhk70529

70530RR_KO1_BamHI ATTGGATCCCGTCCCAAACTGAGATGGAT Forward primer for 5’ flaking sequence of 70530RR

70530RR_KO2_SalI CCGAATGTCGACTACTTTTAACATTTTATTCTCACTTTCC Reverse primer for 5’ flaking sequence of 70530RR

70530RR_KO3_SalI CCGAATGTCGACCTCGAAAAGGAAAACCACTAACTAC Forward primer for 3’ flaking sequence of 70530RR

70530RR_KO4_HindIII ATGAAGCTTAATCACATGCGCAATCAATG Reverse primer for 5’ flaking sequence of 70530RR

70529RR_KO-F TGCCGGGTTCAGAAATAAAA Forward sequencing primer of Δrr70530

70529RR_KO-R TCCGCTGAAAAAGAATAATGG Reverse sequencing primer of Δrr70530

Restriction enzyme sites are indicated in bold

3.2.4 Mouse experiments

Germ-free Swiss Webster mice were bread and maintained in sterile flexible polyurethane isolators at the University of Nebraska Gnotobiotic Mouse Facility. L. reuteri strains (Table 3.1) were tested for production of biofilm according to the protocol described by (Frese et al. 2013).

Briefly, mice were gavaged 100µl with of a PBS solution containing 107 viable L. reuteri cells, then moved into biocontainment cages and housed in groups of two. After 48 hours mice were sacrificed and forestomach tissue and contents collected and processed for culture plating and for microscopic analysis following methods by Frese et al. (2013). Germ-free status of the ventilator system was confirmed by aerobic and anaerobic nonselective (BHI) culture of fecal pellets of two control mice gavaged with sterile PBS and maintained in the same ventilated rack as the experimental mice. All animal experiments were performed in conformity with protocols

#1022 and 1215 approved by the Institutional Animal Care and Use Committee of the University of Nebraska.

3.2.5 In vitro biofilm assays

L. reuteri were grown in mMRS supplemented with 1 g/l sucrose (suMRS) and incubated anaerobically (5% CO2,5% H2, and 90% N2) at 37°C for 48h. Three colonies were then grown overnight in 5 ml suMRS broth before sub-culturing in 5 ml of the same media and grown for another 16h. The cultures were then diluted in fresh suMRS to reach a standardized cell suspension of OD600 0.05 and 10 ml of this suspension were transferred to sterile polystyrene tissue culture plates (SARSTEDT®) and incubated statically at 37 °C for 24 hours.

Quantitative determination of biofilm formation was performed by the spectrophotometric method, which measures the total biofilm biomass, including bacterial cells and the biofilm matrix.

The optical density at 600 nm (OD600) of planktonic cells was determined in the liquid medium which was carefully removed by aspiration without disturbing the adherent biofilm. Next, loosely adhering cells were removed by three gentle washes with 10 ml of sterile phosphate-buffered

saline (PBS). After the washings, the biofilms were harvested by thoroughly scraping the bottom of the culture plates using disposable sterile cell scraper and resuspended in 10 ml of PBS. Biofilm density was determined by measuring the OD600 of this suspension and total numbers of bacteria on the biofilms were determined by the serial dilution plating on suMRS. Each assay was performed at least in triplicate and repeated at least twice.

3.2.6 Mechanical properties of biofilms

Biofilms were grown in tissue culture dishes as described earlier (see in vitro biofilm assays).

The residual culture media with planktonic cells was remove from the plates by decanting and

5 ml of PBS carefully added to the remaining biofilm using an automatic pipette pump set to the lowest dispensing speed. Plates were then agitated on an orbital shaker at 500 rpm for 30s. PBS containing detached biofilm cells matrix was collected by decantation and the OD600 obtained.

The process was repeated at increasing intervals of 30, 60, and 300 seconds.

3.2.7 Confocal Microscopy

Visualization and quantification of biofilm formation in the mouse forestomach was conducted by confocal laser scanning microscopy (CLSM) following methods described previously (Frese et al. 2013). Briefly, forestomach tissues were fixed and bacterial cells stained with propidium iodine.

Confocal images were obtained by a blinded technician from three random tissue sites. Biofilms were quantified by the red-channel pixel area in images using ImageJ (Schindelin et al. 2015)

(Frese et al 2013).

In vitro biofilms for CLSM were grown in ultra-thin, gas-permeable tissue culture plates

(LUMOX®). Attached biofilms were stained with Syto-9 and propidium iodine (FilmTracerTM,

Invitrogen) following manufacturer’s instructions. Five random areas of the biofilm on each plate were scanned and obtained with a Olympus IX-81 spinning disk laser scanning microscope using

Volocity software (PerkinElmer), resulting in 15 imaged areas per sample. The Syto-9 fluorophore

was excited at 525 nm and the emission band-pass filter used was 550 nm. Excitation of propidium iodine was achieved at 620 nm and emission collected using a 650 nm filter.

Simultaneous dual-channel imaging was used to display green and red fluorescence (Hamamatsu

EMCCD camera, 60X/1.42 NA). z-stacks were collected at 1.0-µm intervals and the images were compiled in ImageJ using the Volume Viewer 1.31 plugin.

3.2.8 Scanning Electron Microscopy

L. reuteri wild-type and mutant strains were grown in polystyrene tissue culture dishes as previously described. Biofilms formed at the bottom of the plates were fixed in 0.1 M Sorenson's phosphate buffer containing 2.5% EM grade glutaraldehyde for 48h at room temperature. Fixed biofilms were washed three times for 10 min in 0.1 M PBS buffer (pH 7.4). Samples were then dehydrated with a series of 15 min long washings with solutions at increasing ethanol concentrations (50%, 70%, 90%,100% v/v) Hexamethyldisilazane (HMDS) was introduced using gradually increasing HMDS solutions in ethanol as follows: 75% ethanol-25% HMDS

(Hexamethyldisilazane), 50% ethanol-50% HMDS, 25% ethanol-75% HMDS, and 100% HMDS.

Samples were air dried overnight and then broken down into smaller pieces that were mounted on SEM stubs and sputter coated with Au/Pd (Hummer 6.2 Sputter Coater, Anatech Ltd.). Biofilms were visualized under a scanning electron microscope XL30 (FEI Company) operating at 20 kV.

Pictures were acquired using Scandium 5.0.

3.2.9 Statistical analyses

All statistical analyses were carried out using Graph Pad Prism version 6.2. Statistical significance between the percent biofilm formations of the mutants respective to the WT was determined using a two-sided Mann-Whitney U test. Differences among the optical density values were determined using Student’s t tests.

100

3.3 Results

3.3.1 Genetic organization of the two component system 70529-30

A schematic mapof TCS70529-30 and the surrounding genes are shown in Fig. 3.1. The

hk70529, located at the bp 984225 to 985535 encodes a 436 aa protein containing the Conserved

Protein Domain (COG) 2972 sensor histidine kinase YesM. The rr70530 encodes a protein of 255

aa containing the COG 3279 DNA-binding response regulator of the LytR/AlgR family. Upstream

of the hk70529 are located three small peptides of 32 aa, 53 aa, and 41 aa respectively. Located

downstream of rr70530 is a 52 aa bacteriocin-type signal sequence with a double glycine motif.

Immediately after, two genes encode the C-terminal (461 aa) and N-terminal (299 aa) regions of

an ABC-type bacteriocin transporter. The N-terminal region contains a double-glycine peptidase

domain.

1 2 3 hk70529 rr70530 4 ABC ABC N-terminal

I I I I I I 983805 984225 985535 986310 986986 989136

Figure 3.1 Structural organization of the TCS70529-30 genetic locus Genes were annotated on the IMG ER (http://img.jgi.doe.gov/er). 1, hypothetical protein (99bp); 2, hypothetical protein (162bp); 3 bacteriocin-type signal sequence (126 bp); hk70529, AgrA family, sensor histidine kinase (1311 bp); rr70530, response regulator of the LytR/AlgR family (796 bp); 4, bacteriocin-type signal sequence, contains a double-glycine motif (159 bp); ABC-type bacteriocin transporter (1396 bp); ABC-type bacteriocin transporter, contains an N-terminal double-glycine peptidase domain (720 bp).

3.3.2 Deletion of the rr70530 but not hk70529 the resulted in changes in biofilm in vivo

To characterize the role of TCS 70529-30, in-frame deletions of the individual genes were

generated. Deletion of these genes did not impact the growth of L. reuteri 100-23 in mMRS media

(Fig. 3.2). The ability of mutant strains to colonize and produce biofilms was determined in mono-

associated mice. As shown in Fig. 3.3a, 48 hours postinoculation the biofilm of the hk70529

101

mutant was equivalent to that of the WT. In contrast, deletion of the rr70530 resulted in a 50% reduction in biofilm formation. Interestingly, deletion of these genes did not affect the ability of L. reuteri to colonize the mouse forestomach, as indicated by cell counts (Fig.3.3b). These data suggest that TCS 70529-30 regulates genes that are important for biofilm formation but not for the survivability of L. reuteri in the forestomach.

Figure 3.2 Growth curves of the parent strain L. reuteri 100 -23 and tΔhk70529 and Δrr70530 Strains were grown in MRS medium. Results are shown are the mean ± SD of three independent measurements

102

Figure 3.3 Biofilm formation in the cell counts in the mouse forestomach (a) Quantification of biofilm density relative to biofilm of wild-type strain 100-23 measured as the read pixel area of images obtained by confocal microscopy. Mann Whitney U test ** p<0.01. (b) Cell counts of the wildtype strain and mutants in in the forestomach tissue and content.

3.3.3 Mutation of rr70530 results in biofilm defects in vitro

To investigate the mechanistic basis for the findings observed in vivo, the characteristics of

the biofilms of mutant strains were examined in vitro and compared to the parent strain. Strains

were grown on tissue culture plates and the density of biofilms was determined by optical density.

After removing residual media and planktonic cells, biofilms were washed gently with PBS. The

OD600 readings of the remaining biofilms were consistent with in vivo findings (Fig. 3.4). Deletion

of rr70530, but not hk70529 resulted in significant defects (p<0.05) in biofilm formation. This

analysis also revealed that deletion of the hk70529 increased biofilm formation (Fig. 3.4). In

further agreement with the in vivo results, no differences in cell numbers contained within the

biofilms were observed. Plates were then washed vigorously with PBS, leaving behind only the

most strongly adhered cells and their surrounding biofilm matrix. These extra washing steps

resulted in a significant reduction (p<0.05) of the biofilm density and the cell counts of the

Δrr70530 respective to the WT, whereas no changes in either component were found in the

hk70529 mutant. As shown in Fig. 3.4, the remaining biofilms of the WT and Δrr70530 were

103

reduced almost by half, while the OD600 of the hk70529 was only reduced from an average OD600

5.6 of 4.5. Together these finding supports the notion that differences in biofilm density of the

mutant strains are likely related to structural variations in the biofilm matrix. *

TheFigure graph 3 .represents4 4 In vitro thebiofilm turbidity formation of the biofilms and quantification as reflected byof theirL. reuteri optical 100 density-23 and mutant strains (OD600). Corresponding cell counts are shown. Statistical significant differences from the wildtype are indicated as * p<0.05, ** p<0.01. Mean values of 3 - 6 independent experiments and SDs are presented.

3.3.4 Mutant biofilms exhibit different macroscopic properties and microscale

architectures

Visual inspection of the culture plates reavealed differences in the mutant’s macroscopic

appearance of the biofilms. As shown in Fig. 5a-c, the biofilms produced by the mutant and WT

strains differed in their visible morphology. The WT and Δhk70529 produced coarse biofilms with

defined grooves (Fig. 3.5a and b). In contrast, the biofilm of the rr70530 mutant was smoother

(Fig. 3.5c). A single wash with PBS allowed the visual examination of the internal adhered layers

of the biofilm matrix (Fig. 3.5d – f). Removal of loosely adhered layers of the Δrr70530 biofilm,

104

revealed the presence of grooves or channel like structures. However, these appered to be wider and less defined than those of the WT and the Δhk70529 (Fig. 3.5f).

To further examine the differences underlying the macroscopic differences of the mutant biofilms, we used CLSM to investigate their microscale architecture. Confocal 3D-rendered images of the biofilms are shown in Fig. 3.5 g-i. Consistent with the macroscopic observations, the overall architecture of the biofilm varied amongst strains. Both the WT and Δhk70529 formed compact and well-organized biofilms, but the hk70529 mutant formed a denser biofilm than the parent strain (Fig. 3.5h). On the other hand, the biofilm matrix of the rr70530 mutant was clearly underdeveloped and bacterial cells appeared attached to the bottom of the plates, forming large intercellular gaps.

Scanning electron macroscopic observation showed that biofilms formed by Δhk70529 were more uniform when compared to the WT (Fig. 3.7). On the contrary, large interspersed areas devoid of cells were observed in the biofilms formed by the rr70530 mutant. A closer examination

(x 10 000 magnification), revealed that in some areas the WT and the hk70529 mutant cells were associated with what appears to be an extracellular polymer. Notably these structures although not completely absent, were less frequently observed in association Δrr70530 cells (Fig. 3.7).

Overall, these observations indicate mutations in the TCS70520-30 affect the macroscopic and microscopic characteristics of the biofilms formed by the strain L. reuteri 100-23. These differences in the biofilms formation appear to related presence of exopolymers associated with bacterial cells.

105

Figure 3.5 Macroscopic and microscopic characteristics of in vitro biofilms L .reuteri 100-23 wild type (a, d and g), hk70529 mutant (b, e and h) and rr70530 mutant (c f and i). Top pictures show the biofilm surface after a single PBS wash. Middle pictures show the inner layers of the biofilm after three PBS washes. Bottom pictures are 3D renderings from 1.0-µm z-stacks

106

Figure 3.6 SEM micrographs of L. reuteri 100-23 wild-type and mutant strains on polystyrene surfaces Micrographs were taken at two magnifications, top x 2000, bottom x 10 000. Arrows indicate possible extracellular polymers.

3.3.5 Matrix architecture is associated with changes in biofilm resistance to sheer stress

To obtain a quantitative and physically informative understanding of the observed differences in matrix architecture, and because the presence of channels has been associated with the hydrodynamic properties of biofilms (Wilking et al. 2013), we compared the resistance of the mutant and WT biofilms to hydraulic sheer stress. PBS was added to 24 hour biofilms formed at the bottom of tissue culture plates and resistance to agitation (500 rpm) was measured at increasing time intervals. Resistance was determined relative to the WT as a function of the OD600 of the detached biofilm as a consequence of rotational fluid motion. As shown in Fig. 3.8, differences in biofilm architecture correlate with the ability to resist sheer stress. At all the examined time points the denser biofilm produced by the hk70529 mutant was significantly

(p<0.05) more resilient than the WT and the rr70530. No difference in biofilm strength was found between the WT and the rr70530 during after agitation for 30s and 60s. However, significant

107

differences (p<0.05) were observed at longer times of agitation (120s and 420s). Overall, this analysis revealed that differences in biofilm architecture correlate with resistance to sheer stress.

This data further support the notion that the TCS70529-30 is likely to regulate genes involved in structural properties of the biofilm matrix.

Figure 3.7 Resistance of wildtype and mutant in vitro biofilms to sheer stress

Values are provided in arbitrary units (A.U.) calculated as the reciprocal of the percent OD600 of the detached mutant biofilm respective to that of the wildtype. Means ± standard errors (error bars) of three biological replicates are shown. Values significantly different from the wildtype (p<0.05) are indicated by an asterisk. Statistical comparisons were done using a one-tailed Student’s t test. Significant differences strain are denoted as * p<0.05, ** p<0.01. *** p<0.001

108

3.4 Discussion

In this study, we assessed the role of a rodent-specific TCS in biofilm formation of the rat isolate L. reuteri 100-23. Isogenic deletion mutants where tested for their ability to form biofilms in germ-free mice. Characterization of the physical properties of the biofilms were conducted using a polystyrene plate model. The genetic organization of the TCS70529-30 is that of a typical

TCS where the HK and the RR function together (Capra and Laub 2012). However, deletion of the individual genes resulted in disparate phenotypes. Compared to the WT, deletion of the rr70530 produced a thin and significantly weaker biofilm, while the hk70529 mutant produced a more robust biofilm, highly adherent and more resistant to mechanical breakage.

Although the canonical TCS consists of a signal relay between a histidine kinase and a cognate response regulator, in reality, many of these transduction systems involve multiple components and multiple phosphorelay cascades and feedback loops servings as “checkpoints”.

Moreover, the transcription of one TCS can being controlled by second TCS and certain kinases have shown to interact with non-cognate regulators (Bijlsma and Groisman 2003; Capra and Laub

2012). Biofilm formation in host tissues has proven to be an especially complex process, often involving signal cascades from several TCS. For example, host colonization and biofilm formation of the squid symbiont Vibrio fischeri requires input from at least two sensor kinases (Norsworthy and Visick 2015). Similarly, biofilm formation by the opportunistic pathogen Pseudomonas aeruginosa, also requires cross-talk between HKs (Goodman et al. 2004; Ventre et al. 2006). The rodent strain L. reuteri 100-23 possesses other TCS in addition to 70529-30 (Frese et al. 2011), and at least two of which have been shown to interact at the transcription level (Su and Gänzle

2014). It is therefore quite possible that biofilm formation of L. reuteri 100-23 is controlled by a complex signaling network between TCSs and not by one single system.

According to qualitative and quantitative data obtained in this study, alterations in biofilm formation of mutant strains were related to the structure and strength of the biofilm matrix and not to differences in growth rate or cell numbers within the biofilm. The structural organization and

109

mechanical properties of a biofilm are heavily determined by the composition of its polymeric matrix (Berk et al. 2012; Chew et al. 2014; Persat et al. 2015), which in most bacterial species consists mainly of exopolysaccharides (EPS) proteins, nucleic acids and lipids. EPS mediate cell- to-surface adhesion, cell-cell interactions and serve as scaffolds to the three-dimensional architecture of the biofilm matrix (Flemming and Wingender 2010; Flemming et al. 2016). Mutants unable to synthesize or export EPS are typically deficient in biofilm formation (Karatan and

Watnick 2009). Inactivation of genes involved in EPS production in L. reuteri affects in vitro biofilm formation and ecological performance in the murine gut (Walter et al. 2008). Functional inactivation of signal transduction systems in biofilm-forming bacteria, has shown to affect biofilm formation through mechanisms related to EPS production (Goodman et al. 2004; Ventre et al.

2006; Norsworthy and Visick 2015). For example, deletion of the RetS sensor kinase of

Psudomonas aeruginosa resulted in upregulation of EPS-producing genes, which in turn increased adhesion and biofilm density. Deletion of related HK (ladS) repressing the expression of RetS, decreased both biofilm formation and EPS gene expression (Ventre et al. 2006).

Similarly, Streptococcus mutans defective of the TCS vicRK involved in the transcription of genes involved in EPS biosynthesis, produce aberrant biofilms (Senadheera et al. 2005). Taken together, several studies have shown that the process of biofilm formation is heavily coordinated through TCS and that functional inactivation of these regulatory genes impacts biofilm formation by affecting the transcription of genes involved in EPS production. Although the transcriptional targets of TCS70529-30 are yet unknown, genes involved in EPS production are likely candidates.

Since most experimentally characterized LytTR-containing RRs act as transcriptional activators

(Nikolskaya and Galperin 2002; Galperin 2008), inactivation of rr70530 would impair the transcription of genes involved in EPS synthesis, which would easily explain the biofilm phenotype observed for the rr705030 deficient mutant. Furtheremore, HKs have also been shown to have phosphatase activity, essentially antagonizing the action of RRs (Brooks and Mandel 2016).

Given the increased biofilm phenotype of the hk70529 mutant, the most parcimonous explanation

110

for such obvservation is that hk70529 acts as a phopatase and not a phosphorylase. Deletion of hk70529 would then result in upregulation of genes transcriptionally regulated by rr70530 (e.g

EPS biosynthesis) explaining the phenotypes observed. However, additional experiments, such as those decribed by Brookes and Mendel (2016) are thus warranted to fully elucidate the mechanisms by which TCS70529-30 can be involved in EPS biosynthesis.

Besides EPS, extracellular DNA (eDNA) is another important structural component of the biofilm matrix (Sharma-Kuinkel et al. 2009; Vilain et al. 2009; Harmsen et al. 2010; Svensson,

Pryjma and Gaynor 2014; Okshevsky, Regina and Meyer 2015). The source of the eDNA, is believed to originate from whole-cell lysis or secretion of outer membrane vesicles containing

DNA (Bayles 2007; Sharma-Kuinkel et al. 2009; Hobley et al. 2015). Interestingly, a LytS system known to play a role in cell lysis in S. aureus, was found to be upregulated during biofilm formation of L. reuteri in vivo. It would therefore be warranted to characterize the role of TCS70529-30 in the release of eDNA during biofilms formation of L reuteri. In this sense, it would also be particularly interesting to determine if the expression of LytS is affected by mutations in

TCS70529-30.

In addition to mechanical properties, the structural architecture of the polymeric matrix can have a profound influence in biofilm ecology. Spacing between the bacterial cells effects communication between cells. The organization of the channel network controls the circulation of water and nutrients, as well as the exchange metabolites and signaling molecules that are important for biofilm formation (Wilking et al. 2013; Flemming et al. 2016). Initial attachment is the most critical step for the establishment of biofilms (Flemming and Wingender 2010). Adhesion of

L. reuteri to forestomach is believed to be mediated by a rodent-specific large surface protein

(Frese et al. 2011). Following attachment, a series of adhesive processes link bacteria together into a multilayered three-dimensional structure, ultimately producing a mature biofilm (O’Toole,

Kaplan and Kolter 2000; Nobbs, Lamont and Jenkinson 2009). Partial disintegration of the EPS matrix is also an important step in the biofilm cycle. This process is required for dispersal and

111

colonization of new sites and for the renewal of the bacterial cells within the biofilm (Otto 2013).

Mutations in the arg quorum-sensing system involved in the detachment and dispersal of

S. aureus (Boles and Horswill 2008), resulted in increased adherence and biofilm formation

(Vuong et al. 2000). In this respect, if TCS70529-30 controls biofilm dispersal, deletion of the sensor kinase would halt the detachment process resulting in a denser and more adherent biofilm, explaining the phenotype observed for the hk70529 mutant. Real time microscopy imaging paralleled trascriptome profiling have been proven useful tools to analyse the dynamics of the biofilm development and detachment (Nicholson, Conover and Deora 2012). Similar experiments are thus warrante to determine de involvement of TCS70529-30 in biofilm dispersal.

In conclusion, this work provides empirical evidence of a rodent specific signal transduction system playing a role in biofilm formation of L. reuteri 100-23. However, its precise function and the transcriptional genetic targets remain to be identified. Since deletion of individual genes did not affect biomass accumulation but rather exerted effects on biofilm architecture, genes related to the development of the biofilm matrix (e.g. EPS production and release of eDNA) seem to be the most likely candidates, though further experimental evidence will be required to confirm such hypotheses.

The ecological fate of L. reuteri depends on its ability to occupy the forestomach niche and persist inside the murine host. It is therefore highly unlikely for the entire process of biofilm formation to be devoted to one single signal transduction system. Instead, the mechanisms governing biofilm are expected to involve complex signaling cascades, likely orchestrated through multiple system transduction systems; as it has been shown for other symbionts (Norsworthy and

Visick 2015). Deciphering the role of TCS70530-29 in biofilm formation can foster our understanding of the molecular mechanisms involved in the adaptation of L. reuteri to the vertebrate gut. Important challenges for the future studies are therefore, to identify the signals

112

recognized by the sensor kinase hk70529 and to determine the target genes of the rr70530.

Findings from this work open avenues for such studies.

3.5 References

Abbas Hilmi HT, Surakka A, Apajalahti J et al. Identification of the Most Abundant

Lactobacillus Species in the Crop of 1- and 5-Week-Old Broiler Chickens. Appl Environ Microbiol

2007;73:7867–73.

Abee T, Kovács AT, Kuipers OP et al. Biofilm formation and dispersal in Gram-positive bacteria. Curr Opin Biotechnol 2011;22:172–9.

Alm E, Huang K, Arkin A. The evolution of two-component systems in bacteria reveals different strategies for niche adaptation. PLoS Comput Biol 2006;2:1329–42.

Altermann E. Tracing Lifestyle Adaptation in Prokaryotic Genomes. Frontiers in Microbiology

2012;3:1-17

Bayles KW. The biological role of death and lysis in biofilm development. Nat Rev Microbiol

2007;5:721–6.

Brookes and Mendel.The histidine kinase BinK is a negative regulator of biofilm formation and squid colonization. J Bacteriol 2016 198:2596-2507

Bijlsma JJE, Groisman E a. Making informed decisions: regulatory interactions between two- component systems. Trends Microbiol 2003;11:359–66.

Boles BR, Horswill AR. agr-mediated dispersal of Staphylococcus aureus biofilms. PLoS

Pathog 2008;4, DOI: 10.1371/journal.ppat.1000052.

Branda SS, Vik S, Friedman L et al. Biofilms: the matrix revisited. Trends Microbiol

2005;13:20–6.

Capra EJ, Laub MT. Evolution of two-component signal transduction systems. Annu Rev

Microbiol 2012;66:325–47.

113

Davey ME, O’toole GA. Microbial Biofilms: from Ecology to Molecular Genetics. Microbiol Mol

Biol Rev 2000;64:847–67.

Duar RM, Frese SA, Fernando SC et al. Experimental determination of host adaptation of

Lactobacillus reuteri to different vertebrate species. Appl Environ Microbiol 2017:AEM.00132-

17.(Chapter two of this thesis)

Flemming H-C, Wingender J, Szewzyk U et al. Biofilms: an emergent form of bacterial life.

Nat Rev Microbiol 2016;14:563–75.

Flemming H, Wingender J. The biofilm matrix. Nat Rev Microbiol 2010;8:623–33.

Frese SA, Benson AK, Tannock GW et al. The evolution of host specialization in the vertebrate gut symbiont Lactobacillus reuteri. PLoS Genet 2011;7:16.

Frese SA, Mackenzie DA, Peterson DA et al. Molecular characterization of host-specific biofilm formation in a vertebrate gut symbiont. PLoS Genet 2013;9:e1004057.

Galperin MY. Telling Bacteria: Do Not LytTR. Structure 2008;16:657–9.

Goodman AL, Kulasekara B, Rietsch A et al. A signaling network reciprocally regulates genes associated with acute infection and chronic persistence in Pseudomonas aeruginosa. Dev Cell

2004;7:745–54.

Harmsen M, Lappann M, Knøchel S et al. Role of extracellular DNA during biofilm formation by Listeria monocytogenes. Appl Environ Microbiol 2010;76:2271–9.

Hobley L, Harkins C, MacPhee CE et al. Giving structure to the biofilm matrix: an overview of individual strategies and emerging common themes. FEMS Microbiol Rev 2015;39:649–69.

Krumbeck JA, Marsteller NL, Frese SA et al. Characterization of the ecological role of genes mediating acid resistance in Lactobacillus reuteri during colonization of the gastrointestinal tract.

Environ Microbiol 2016;18:2172–84.

Leser TD, Amenuvor JZ, Jensen TK et al. Culture-Independent Analysis of Gut Bacteria: the

Pig Gastrointestinal Tract Microbiota Revisited. Appl Environ Microbiol 2002;68:673–90.

114

Li Y-H, Lau PCY, Tang N et al. Novel Two-Component Regulatory System Involved in Biofilm

Formation and Acid Resistance in Streptococcus mutans. J Bacteriol 2002;184:6333–42.

Mandel MJ, Wollenberg MS, Stabb E V et al. A single regulatory gene is sufficient to alter bacterial host range. Nature 2009;458:215–8.

Markowitz VM, Mavromatis K, Ivanova NN et al. IMG ER: a system for microbial genome annotation expert review and curation. Bioinformatics 2009;25:2271–8.

Mikkelsen H, Sivaneson M, Filloux A. Key two-component regulatory systems that control biofilm formation in Pseudomonas aeruginosa. Environ Microbiol 2011;13:1666–81.

Monedero V, Revilla-Guarinos A, Zúñiga M. Physiological Role of Two-Component Signal

Transduction Systems in Food-Associated Lactic Acid Bacteria. Adv Appl Microbiol 2017, DOI:

10.1016/bs.aambs.2016.12.002.

Nicholson TL, Conover MS, Deora R. Transcriptome profiling reveals stage-specific production and requirement of flagella during biofilm development in Bordetella bronchiseptica.

PLoS One 2012;7, DOI: 10.1371/journal.pone.0049166.

Nikolskaya AN, Galperin MY. A novel type of conserved DNA-binding domain in the transcriptional regulators of the AlgR/AgrA/LytR family. Nucleic Acids Res 2002;30:2453–9.

Nobbs AH, Lamont RJ, Jenkinson HF. Streptococcus Adherence and Colonization. Microbiol

Mol Biol Rev 2009;73:407–50.

Norsworthy AN, Visick KL. Signaling between two interacting sensor kinases promotes biofilms and colonization by a bacterial symbiont. Mol Microbiol 2015;96:233–48.

O’Toole G, Kaplan HB, Kolter R. Biofilm Formation as Microbial Development. Annu Rev

Microbiol 2000;54:49–79.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Okshevsky M, Regina VR, Meyer RL. Extracellular DNA as a target for biofilm control. Curr

Opin Biotechnol 2015;33:73–80.

115

Otto M. Staphylococcal infections: mechanisms of biofilm maturation and detachment as critical determinants of pathogenicity. Annu Rev Med 2013;64:175–88.

Perez-Casal J, Price JA, Maguin E et al. An M protein with a single C repeat prevents phagocytosis of Streptococcus pyogenes: use of a temperature-sensitive shuttle vector to deliver homologous sequences to the chromosome of S. pyogenes. Mol Microbiol 1993;8:809–19.

Sambrook J, W Russell D. Molecular Cloning: A Laboratory Manual. Cold Spring Harb Lab

Press Cold Spring Harb NY 2001:999.

Schindelin J, Rueden CT, Hiner MC et al. The ImageJ ecosystem: An open platform for biomedical image analysis. Mol Reprod Dev 2015;82:518–29.

Senadheera MD, Guggenheim B, Spatafora GA et al. A VicRK signal transduction system in

Streptococcus mutans affects gtfBCD, gbpB, and ftf expression, biofilm formation, and genetic competence development. J Bacteriol 2005;187:4064–76.

Sharma-Kuinkel BK, Mann EE, Ahn J-S et al. The Staphylococcus aureus LytSR two- component regulatory system affects biofilm formation. J Bacteriol 2009;191:4767–75.

Su MS, Schlicht S, Gänzle MG. Contribution of glutamate decarboxylase in Lactobacillus reuteri to acid resistance and persistence in sourdough fermentation. Microb Cell Fact 2011;10

Suppl 1:S8.

Su MSW, Gänzle MG. Novel two-component regulatory systems play a role in biofilm formation of Lactobacillus reuteri rodent isolate 100-23. Microbiology 2014:4–10.

Svensson SL, Pryjma M, Gaynor EC. Flagella-mediated adhesion and extracellular DNA release contribute to biofilm formation and stress tolerance of Campylobacter jejuni. PLoS One

2014;9:e106063.

Tannock GW. The lactic microflora of pigs, mice and rats. The lactic acid bacteria volume 1.

Boston, MA, MA: Springer US, 1992, 21–48.

Toft C, Andersson SGE. Evolutionary microbial genomics: insights into bacterial host adaptation. Nat Rev Genet 2010;11:465–75.

116

Ventre I, Goodman AL, Vallet-Gely I et al. Multiple sensors control reciprocal expression of

Pseudomonas aeruginosa regulatory RNA and virulence genes. Proc Natl Acad Sci U S A

2006;103:171–6.

Vilain S, Pretorius JM, Theron J et al. DNA as an adhesin: Bacillus cereus requires extracellular DNA to form biofilms. Appl Environ Microbiol 2009;75:2861–8.

Vuong C, Saenz HL, Götz F et al. Impact of the agr quorum‐sensing system on adherence to polystyrene in staphylococcus aureus. J Infect Dis 2000;182:1688–93.

Walter J, Britton RA, Roos S. Host-microbial symbiosis in the vertebrate gastrointestinal tract and the Lactobacillus reuteri paradigm. Proc Natl Acad Sci U S A 2011;108 Suppl:4645–52.

Walter J, Heng NCK, Hammes WP et al. Identification of Lactobacillus reuteri Genes

Specifically Induced in the Mouse Gastrointestinal Tract. Appl Environ Microbiol 2003;69:2044–

51.

Walter J, Schwab C, Loach DM et al. Glucosyltransferase A (GtfA) and inulosucrase (Inu) of

Lactobacillus reuteri TMW1.106 contribute to cell aggregation, in vitro biofilm formation, and colonization of the mouse gastrointestinal tract. Microbiology 2008;154:72–80.

Wesney E, Tannock GW. Association of rat, pig, and fowl biotypes of lactobacilli with the stomach of gnotobiotic mice. Microb Ecol 1979;5:35–42.

Wilking JN, Zaburdaev V, De Volder M et al. Liquid transport facilitated by channels in Bacillus subtilis biofilms. Proc Natl Acad Sci U S A 2013;110:848–52.

Wilson CM, Loach D, Lawley B et al. Lactobacillus reuteri 100-23 Modulates Urea Hydrolysis in the Murine Stomach. Appl Environ Microbiol 2014;80:6104–13.

Yanisch-Perron C, Vieira J, Messing J. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 1985;33:103–19.

Zhang Y, Lei Y, Khammanivong A et al. Identification of a novel two-component system in

Streptococcus gordonii V288 involved in biofilm formation. Infect Immun 2004;72:3489–94.

117

4. Chapter four: Metabolic cross-feeding between 1,2 propanediol-producing intestinal bacteria and the human isolate Lactobacillus reuteri ATCC PTA6475

118

4.1 Introduction

The human intestine harbors a diverse and complex microbial community composed of trillions of metabolically active bacteria. This community of microbes exists in an intricate mutualistic symbiosis with the host, where microbes provide a number of key functions to the host’s well-being and fitness and in return receive secure habitats and nutrients (Bäckhed et al.

2005; Clemente et al. 2012).

The main energy sources that support the microbial community in the human intestine are nondigestible dietary compounds and endogenous host secretions such as mucin (Koropatkin,

Cameron and Martens 2012). As is typical in complex microbial ecosystems, much of the nutrient acquisition in the gut depends on a dynamic network of competitive and cooperative interrelationships amongst members of the community (Coyte, Schluter and Foster 2015). One such interaction, referred to as cross-feeding, occurs when compounds derived from the metabolic output of one bacterium serve as metabolic input for another (Seth and Taga 2014).

Cross-feeding of nutrients is central theme in the fermentative degradation of complex carbohydrates and has a huge impact on the final balance of short chain fatty acids produced in the gut (Flint et al. 2012). One important route for the production of butyrate occurs though metabolic cross-feeding interactions between lactate-producing species and lactate-utilizing butyrogenic species (Duncan, Louis and Flint 2004; Rivière et al. 2015). Another mechanism involves cross-feeding of fractions released from the partial breakdown of complex polymers

(Belenguer et al. 2006), such as the degradation of fructans, starch, xylan, and mucin polymers which have been shown to require a series of syntrophic interactions involving different species with complementary enzymatic repertoires and nutritional requirements (Egan et al. 2014a,

2014b; Rakoff-Nahoum, Coyne and Comstock 2014; Turroni et al. 2015).

A prerequisite to enter such intricate nutrient networks is the specialization towards a subset of the available substrates, as is often apparent in the genomes of human gut symbionts (Xu et al. 2003; Sela et al. 2008; Martens et al. 2011). Lactobacillus reuteri is a host-specific gut symbiont

119

to different vertebrate species (Oh et al. 2010; Frese et al. 2011; Walter, Britton and Roos 2011;

Duar et al. 2017) (Chapter 2) and one of the few lactobacilli regarded as autochthonous to the human gut (Reuter 2001). Comparisons between the genomes of human and rodent isolates revealed that strains have adapted to the gut conditions of each corresponding host (Frese et al.

2011). In rodents, L reuteri colonizes by producing host-specific biofilms on the keratinized epithelium of the forestomach. Here it obtains simple carbohydrates and growth factors from the host’s diet. (Frese et al. 2013). Analogous non-secretory regions, have not been described in the human gastrointestinal tract where L. reuteri is likely to be restricted to a planktonic lifestyle

(Walter, Britton and Roos 2011). This lifestyle poses important challenges for nutrient acquisition.

L reuteri, like other lactobacilli, relies on the availability of fermentable sugars for growth but such sugars rarely reach the colon. Therefore, how L. reuteri satisfies its growth requirements in the human colon is currently unknown. One interesting finding from genome comparative analyses was that genes involved adhesion and biofilms formation, presumably superfluous to a lifestyle in the human colon, have been largely deleted from the genomes of human strains. However, a 58- gene cluster encoding several biosynthetic enzymes has been conserved in human isolates and mostly absent from the genomes of animal-associated strains (Frese et al. 2011).

Encoded in the pdu-cbi-cob-hem cluster (pdu-cluster, for short) are enzymes involved in the biosynthesis of cobalamin and the broad spectrum antibiotic reuterin (Morita et al. 2008; Sriramulu et al. 2008). Also in this cluster is the is the glycerol/diol dehydratase PduCDE involved in the utilization of glycerol and 1,2 propanediol (1,2 PD). Metabolism of glycerol improves the competitiveness of L. reuteri in sourdough and addition of 1,2 PD increases cell yield of L. reuteri growing in galactooligosaccharides (Rattanaprasert et al. 2014). In the human intestine, deoxy hexoses, namely rhamnose and fucose, are readily available as part plant-derived dietary fibers and host mucins, respectively. L reuteri cannot utilize these sugars directly (Rattanaprasert,

2014). However, a number of intestinal bacteria metabolize fucose and rhamnose into 1,2 PD

(Saxena et al. 2010; Reichardt et al. 2014). Hence, the ability of L. reuteri to metabolize 1,2PD

120

has been hypothesized to be relevant to its ecological performance in the human gut (Walter,

Britton and Roos 2011). The primary goal of this project was to determine the role of the pdu- cluster in the context of metabolic cross-feeding of 1,2 PD. For this purpose we selected strains from two different 1,2 PD-producing species that are commonly found in the human gut

(Escherichia coli and Bifidobacterium breve) and determined the ability of L. reuteri to benefit from 1,2 PD provided via cross-feeding through the fermentation of rhamnose and fucose by other species.

4.2 Materials and Methods

4.2.1 Bacterial strains and media

Bacterial strains used in this study are listed in Table 1. L. reuteri strains were grown in MRS medium (Difco) under anaerobic conditions (5% CO2, 5% H2, and 90% N). B. breve strains were grown anerobically in Reinforced Clostridial Medium broth (Oxoid Ltd). E. coli were grown in Luria-

Bertani (LB) broth with agitation. All cultivations were at 37 °C.

Table 4.1 Bacterial strains used in this study Strain (other names) Origen Relevant characteristics Reference or Source Bifidobacterium breve Stool of a nursing L-fucose utilizer, 1,2 PD (Mazé et al. 2007) UCC2003 infant producer Escherichia coli L-rhamnose utilizer, 1,2 PD The Coli Genetic Stock Derived lab strain MG1655 producer Center (CGSC) Lactobacillus reuteri ATCC PTA 6475 Breast milk Wild type, 1,2 PD utilizer BioGaia AB (MM4-1A) Lactobacillus reuteri Derivative of Deletion mutant, glycerol/diol This study ΔpduCDE ATCC PTA 6475 dehydratase PduCDE

121

4.2.2 Generation of the L. reuteri ΔpduCDE mutant

A suicide shuttle vector for pduCDE deletion (pVPL3478) was constructed by Ligase

Cycling Reaction (LCR) as described in de Kok et al. (2014). Up-stream flanking (oVPL1335 and oVPL1336, 847bp) and down-stream flanking region (oVPL1337 and oVPL1338, 945bp) from L. reuteri ATCC PTA 6475 pduCDE locus were amplified by PCR. Backbone of suicide shuttle vector pVPL3002 (3.7kb pORI19_EmR derivative with vancomycin counter-selection marker) was amplified with oVPL187-188. Phusion Hot Start II DNA polymerase (Thermo Fisher Scientific) by

PCR. Amplicons were purified by using Gen-jet PCR-purification kit (Thermo Fisher Scienctific) and vector backbone was subsequently treated with DpnI (Fast Digest, Thermo Fisher Scientific) followed by PCR clean-up. 4 nM of each PCR amplicons were mixed and phosphorylated by using poly nucleotides kinase (Thermo Fisher Scientific) followed by EtOH precipitation. LCR was conducted in PCR machine as described in a previous study (Kok et al. 2014) and three bridging oligos (oVPL1339, 1340, and 1341) were used for LCR. LCR ligates were pellet-paint® precipitated followed by E. coli EC1000 transformation. 5 μg total plasmid DNA (pVPL3478) was used to transform L. reuteri ATCC PTA 6475. L. reuteri electro-competent cell preparation and electroporation were performed following protocols described in Oh and Van Pijkeren (2014).

Single crossover was screened by PCR from colonies on MRS containing 5 μg/ml erythromycin and subsequently ΔpduCDE mutant colonies derived from two passages in plain MRS broth were screened on a MRS plate containing 500 μg/ml vancomycin followed by colony PCR (oVPL1342 and oVPL1344) and sequencing analysis (Table S1).

4.2.3 Basal media for fermentations

The basal MRS (bMRS) was devised based on a recipe described elsewhere (Stolz et al.

1995). The bMRS media consisted of (per liter): 5g of peptone, 2.5g of beef extract, 2.5g of yeast extract, 1.5g of ammonium chloride, 1.6g of monopotassium phosphate, 2g of dipotassium phosphate, 0.025 g of magnesium sulfate, 0.0175g of mangase sulfate, 0.25g of L-cysteine

122

hydrochloride and 0.5 ml of tween 80. Basal media was autoclaved for 15 min at 120°C and the pH adjusted to 6.6.

4.2.4 Screening of 1.2 PD utilization by L reuteri.

Utilization of 1,2 PD and the metabolites produced during co-fermentation with glucose were determined for the wild-type L. reuteri ATCC PTA 6475 and a pduCDE mutant strain (Table 1).

Cells from overnight cultures were harvested by centrifugation and washed in sterile PBS (pH 7.0) as previously described (Rattanaprasert et al. 2014). Washed cells were suspended in bMRS and inoculated at 1 % into 25 ml of bMRS media containing 20 mM of 1,2 PD alone, 50 mM of glucose alone, or a combination of both. Cell growth was monitored by optical density at 600 nm (OD600).

All experiments were performed in triplicate at 37°C under anoxic conditions.

4.2.5 In vitro fermentations for 1,2 PD production

Carbohydrates were added separately to the bMRS from filter-sterilized stock solutions as summarized in Table 2. L- rhamnose was added to a final concentration of 40 mM and L-fucose was added to reach 20 mM. Media were left overnight to reduce in the anaerobic chamber. The following morning 200 ml of bMRS containing 40 mM L-rhamnose was inoculated at 1% with E. coli grown overnight in LB. B. breve grown overnight in MRS (Difco) was added at 1% inoculum to 200 ml of bMRS media containing 20 mM of fucose. Fermentations were conducted for 24 h at

37 °C under anaerobic conditions. Bacterial cells were then removed by centrifugation (5 000 × g for 5 minutes). Both media were resuplemented (using solid ingredients) to regain the nutrient and salts concentration of the bMRS, filtered sterilized (0.2 µm), and stored at 4 °C. Cross-feeding assays were conducted within 2 days.

123

4.2.6 Cross-feeding assays

Growth of L. reuteri in media previously fermented by E. coli was tested with and without the addition of glucose (25 mM) as summarized in Table 4.2. Because growth of B. breve was not achieved in bMRS media, the effect of fucose on the growth of L. reuteri was determined in bMRS media containing fucose (20 Mm) in the presence and absence of glucose (Table 4.2). L. reuteri inoculums for cross-feeding assays were prepared as described previously (see Methods sections “Screening for 1,2 PD utilization by L. reuteri”). Cell growth was monitored for 12 hours.

OD600 readings and 1 ml samples for HPLC were collected every 3 hours. All experiments were performed in triplicate at 37°C under anoxic conditions.

4.2.7 Analytical methods

The amount of 1,2 PD produced from L-frucose and L-rhamnose and present in the resuplemented media, as well as the production of lactate, acetate, ethanol, propanol and propionate and residual amounts of 1,2PD after fermentations were determined by HPLC (Lin and Gänzle 2014). During the cross-feeding growth assays 1 ml samples were taken for metabolites analysis at 0, 6, and 12 hours.

4.2.8 Statistical analysis

Significant differences in OD600 values were determined by Student’s t-tests calculated using

GraphPad Prism version 6.07

124

Table 4.2 List of media used for cross-feeding experiments

Before fermentation Reconditioned for cross-feeding Approx. 1,2 Added Fermenting Deoxy Added PD deoxy Abbreviationb Purpose organism hexose glucose produced a hexose

E. coli Determine the growth rate of L. reuteri in a none not detected 25 mM none Eco + (Glc) resuplemented media containing glucose Determine if L. reuteri can use rhamnose as a none not detected none Rha(40mM) Eco + (Rha) growth substrate Determine the effect of rhamnose in co- none not detected 25 mM Rha(40mM) Eco+(Rha+Glc) fermentation glucose on the growth rate of L. reuteri Determine if L. reuteri can use 1,2 (produced Rha 40mM 15mM none none Eco_1,2PD from Rha) as a growth substrate Determine the effect of 1,2 PD (produced from Rha 40mM 15mM 25 mM none Eco_1,2PD+(Glc) Rha) in co-fermentation with glucose, in the growth rate of L. reuteri B. breve

125 Determine effect of 1,2 PD (produced from Fuc) Fuc 20mM 15mM 25 mM none Bre_1,2PD+(Glc) in co-fermentation glucose

Other media Determine if L. reuteri can use fucose as a growth none not detected none Fuc(20mM) bMRS+Fuc substrate

Determine the effect of fucose in co-fermentation 25 mM Fuc(20mM) bMRS+Fuc+Glc with glucose on the growth rate of L. reuteri

aas determined by HPLC. bsugars added during the reconditioning step are shown in brackets Eco: E. coli; Bre: B. breve Rha: rhamnose, Fuc: fucose, Glc: glucose

4.3 Results

4.3.1 Effect of 1,2 PD on growth of L. reuteri

Growth of L. reuteri ATCC PTA 6475 wild type and a pduCDE mutant in bMRS containing

glucose was determined by OD and compared to growth in this same media supplemented with

1,2 PD. Growth patterns, utilization of 1,2 PD and its effects on glucose metabolism were

determined under anaerobic conditions and monitored for 12 hours (Fig 1 a-c). When grown in

bMRS with glucose (bMRS+Glc), L. reuteri wild type and the pduCDE mutant reached comparable

ODs (Fig1a). However, when 1,2 PD was added to this media (bMRS+glc+1,2PD) the wild type

strain grew faster and reached a higher OD, whereas no change in growth was observed for the

pduCDE mutant (Fig 1b). As shown in Fig. 1c, addition of only 1,2 PD to bMRS media resulted in

no growth for either strain, confirming that L. reuteri are unable to utilize 1,2 PD as a sole carbon

source.

Figure 4.1 Growth characteristics of L. reuteri ATCC PTA 6475 and ΔpduCDE

Growth of L. reuteri ATCC PTA 6475 and ATCC PTA 6475 훥 pduCDE was monitored by OD600 for 12 hours in bMRS containing (a) glucose (50mM),(b) glucose (50mM) and 1,2 PD (20mM) (c) 1,2 PD (20mM). Results presented are mean values obtained from three separate experiments. Error bars represent standard deviations.

126

At the end of the fermentation, the concentration of 1,2 PD, propanol and propionate was measured in the culture supernatants by HPLC. As shown in Table 3, almost all the 1,2 PD added to the media was metabolized into propanol and propionate by the wild type strain. With propanol being produced more abundantly than propionate. On the other hand, after 12 h of fermentation with the pduCDE mutant, the residual amount of 1,2 PD was only 2mM less than the amount initially added to the media. Production of propanol or propionate was not detected in the culture supernatant of the pduCDE mutant strain.

Lactate, acetate and ethanol were also measured in the culture supernatants to determine the effects of 1,2 PD in the metabolism of glucose (Table 3). When 1,2 PD was added to the media, production of ethanol decreased resulting in a concomitant increase of acetate, while production of lactate remained constant. The metabolic end products of glucose fermentation of the pduCDE mutant strain were similar to those of the wild type and remained at the same molar ratio independent of the addition of 1,2 PD. Overall, these results indicate that simultaneous fermentation of glucose and 1,2 PD increases the conversion of glucose into acetate, resulting in approximate equimolar amounts of acetate and ethanol, and these effects are abrogated with the inactivation of the glycerol/diol dehydratase pduCDE

4.3.2 Production of 1,2 PD from deoxy hexoses

We then sought to determine whether 1,2 PD excreted from a “producer” species could be utilized by L. reuteri. For the selection of producer strains we established four criteria (i) the organism must be commonly found in the human gastrointestinal tract (GIT) (ii) be nonpathogenic,

(iii) able to produce 1,2 PD from deoxy hexoses commonly present in the GIT under anaerobic conditions, and (iv) be able to grow on bMRS media. This last criteria would come to importance in future co-culture experiments of the 1,2 PD-producer and L. reuteri. Based on the above criteria we selected E. coli and B. breve as the 1,2 PD-producer organisms. Strains of E coli can utilize

L-rhamnose, which is available in the gut as it is present in food-derived pectic and hemicellulosic

127

polysaccharides and, under anaerobic conditions, excrete 1,2 PD into the medium (Baldomà and

Aguilar 1988). Strains of B. breve can utilize L-fucose, which is a component of host-derived glycans, and although production of 1,2 PD from L-fucose had to our knowledge not been demonstrated experimentally, most B. breve strains possess the genetic machinery to produce

1,2 PD (Egan et al. 2014a; Bunesova, Lacroix and Schwab 2016). For the production of 1,2 PD,

E. coli and B. breve were grown anaerobically in bMRS containing different amounts of L- rhamnose or L-fucose, respectively. The highest yield of 1,2 PD was determined to come from fermentation of 40mM of L-rhamnose and 20mM of L-fucose.

128

Table 4.3 Effect of addition of 1,2 PD on the end products of heterolactic fermentation of glucose by L. reuteri

Metabolites produced in the Metabolites produced from

presence/absence of 20mM of 1,2 PD fermentation of 55 mM glucose 1,2 PD Approx. molar ratio Strain Media Propanol Propionate Lactate Ethanol (E) Acetate (A) consumed E:A

Wild-type bMRS - - - 33 ± 3.8 18.1 ± 4.8 1.0 ± 0.2 18:1 bMRS+1,2PD -16.5+0.2 14.6 ± 2.9 2.7 ± 0.1 29.9 ± 1.2 9.7 ± 1.9 7.4 ± 0.2 1:1 ΔpduCDE bMRS - - - 33.7 ± 0.5 20.3 ± 0.4 0.94± 0.00 20:1 bMRS+1,2PD -2.1 ± 1.3 n.d n.d 26.4 ± 1.2 13.8 + 1.8 0.80 ± 0.08 17:1

129 n.d, not detected

4.3.3 Cross-feeding of E. coli-produced 1,2 PD

The next step was to determine whether L. reuteri could grow on a filter sterilized reconditioned medium derived from growth of a 1,2 PD-producing species. As shown in Fig 4.2a, both the L. reuteri ATCC 6475 and the pduCDE mutant are able to grow on an E. coli reconditioned media containing glucose [Eco+(Glc)] and reached comparable ODs. We also determined the effects of rhamnose on the growth of L. reuteri. As shown in Fig 4.2b, L reuteri strains were unable to grow when the reconditioned media contained only rhamnose [Eco+(Rha)].

Furthermore, addition of rhamnose had no effect in the growth of L. reuteri in glucose

[Eco+(Rha+Glc)] (Fig 4.2c) as the highest OD reached was equivalent in the absence (wild-type

2.6 ± 0.1, ΔpduCDE 2.5 ± 0.1) or presence (wild-type 2.4 ± 0.1, ΔpduCDE 2.3 ± 0.1) of rhamnose.

Finally, we confirmed that neither the wild type or the mutant strain were able to grow on reconstituted media containing 1,2 PD produced from the fermentation of L-rhamnose as the sole substrate (Fig 4.2d).

Having confirmed that L. reuteri could grow on an E. coli-reconditioned media and that L- rhamnose had no effect on growth, we continued with the cross-feeding experiments. The growth of L reuteri wild-type and the pduCDE mutant were monitored for 12 hours in a reconditioned media containing E.coli-produced 1,2 PD and 25mM of glucose [Eco_1,2PD +(Glc)]. As shown in Fig. 3, both mutant and wild type strains grew to the same OD during the first 6 hours. However at the 9 hour time point the wild-type strain reached a significantly higher (p<0.05) OD than the

ΔpduCDE strain. This difference in OD was also observed at the 12 hour time point (Fig 3a).

Analysis of the 1,2 PD, propanol and propionate in culture supernatants revealed that in co- fermentation with glucose, the wild-type strain metabolized 1,2 PD into propanol and propionate, whereas the pduCDE mutant was unable to metabolize 1,2 PD (Fig.3 b-d) Overall these analyses confirm that L. reuteri can cross-feed from 1,2 PD produced by the fermentation of L-rhamnose by E. coli.

130

Figure 4.2 Growth characteristics of L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE (triangles) in a reconditioned E. coli spent medium

Growth of L. reuteri ATCC PTA 6475 and the PTA 6475 the pduCDE mutant was monitored by OD for 12 hours in spent medium in which E. coli has been previously grown and reconditioned to contain (a) glucose (25mM), (b) rhamnose (40mM) (c) rhamnose and glucose (d) only 1,2 PD produced from the fermentation of L-rhamnose. Results presented are mean values obtained from three separate experiments. Error bars represent standard deviations.

131

Figure 4.3 Growth and metabolites produced by L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE (triangles) growing in a reconditioned E. coli spent medium

(a)Growth as determined by OD600 (b) utilization of 1,2 PD, and production of (c) propanol and (d) propionate. Results presented are mean values obtained from three separate experiments. Error bars represent standard deviations. Significant differences (p<0.05) in OD values are denoted with an asterisk

132

4.3.4 Cross-feeding of B. breve-produced 1,2 PD

First we sought to determine if addition of L-fucose had any effects on the growth of L. reuteri.

As shown in Fig. 4a, L. reuteri strains were unable to grow on bMRS supplemented with L-fucose

alone (bMRS+Fuc). Additionally, both wildtype and mutant grew to comparable ODs in bMRS

containing both glucose and L-fucose (bMRS+Glc+Fuc) (Fig 4b). Furthermore, neither L. reuteri

strain was able to grow in in a reconditioned B. breve media containing only 1,2 PD produced

form L-fucose (Fig. 4c). Based on these findings we concluded that L-fucose had no effects in the

growth of L reuteri and confirmed that strains cannot use 1,2 PD as carbon source. We then

continued with the cross-feeding experiments.

Figure 4.4 Growth characteristics of L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE (triangles) bMRS and in a reconditioned B. breve spent medium

Growth in was monitored by OD600 for 12 hours in (a) bMRS containing fucose (b)bMRS containing glucose and fucose, and in a (c) a reconditioned B. breve medium containing 1,2 PD from fermentation of L-fucose. Results presented are mean values obtained from three separate experiments. Error bars represent standard deviations.

As shown in Fig. 5 different growth profiles were observed when glucose was added to the

reconditioned media containing 1,2 PD produced from the fermentation of L-fucose

[Bre_1,2PD+(Glc)]. During the first 6 hours of growth on this media, both mutant and the wild-type

strains reached equivalent ODs (Fig 5 a). However, after 9 hours the mutant appeared to have

reached stationary while the wild-type strain continued growing, reaching a significantly higher

(p<0.05) ODs at the 9 hour time point and the 12 hour time point (Fig 5 a). Quantitative analysis

of 1,2 PD, propanol and propionate showed that 1,2 PD was metabolized into propanol and

133

propionate by the wild type strain (Fig 5. b-d). Conversely, the pduCDE mutant was not capable of utilizing 1,2 PD as evidenced by the residual amount of 1,2 PD in the supernatant (Fig 5. b).

Taken together these findings show that L. reuteri can cross-feed from 1,2 PD produced from the fermentation from deoxy hexoses by two common GIT inhabitants, E. coli and B.breve. In co- fermentation with glucose, disproportionation of 1,2 PD into propanol and propionate increases the growth of L reuteri was determined by OD. Inactivation of the glycerol/dehydratase pduCDE abrogates such effects.

Figure 4.5 Growth and metabolites produce by L. reuteri ATCC PTA 6475 (circles) and PTA 6475 ΔpduCDE (triangles) growing in a reconditioned B. breve spent medium

134

4.4 Discussion

Substrate breakdown by intestinal bacteria involves the combined metabolic action of primary fermenters and syntrophs that consume generated byproducts (Fischbach and Sonnenburg 2011;

Koropatkin, Cameron and Martens 2012). Here we demonstrate that trophic interaction of metabolites can occur between human isolates of L. reuteri and strains from two common intestinal species, E. coli and B. breve. We further demonstrate that this interaction is facilitated by the pdu cluster, and more specifically, the propanediol dehydratase PduCDE, which allows L. reuteri to cross feed on fermentation end products (1,2 PD) of rhamnose and fucose.

One major determinant to the success of a bacterium in a particular niche depends largely on its ability to obtain resources to generate energy. L. reuteri, like other heterofermentative lactobacilli, generates ATP from carbohydrates via the phosphoketolase pathway (PKP) (Arskold et al. 2008). In the absence of electron acceptors to regenerate reduced cofactors, the PKP is energetically inefficient (Gänzle 2015). In the mouse forestomach, L. reuteri obtains fermentable sugars and electron acceptors (e.g. fructose) from the digesta (Tannok 2012). These compounds are much harder to come across in the distal portions of the human intestine, as they become absorbed early in the digestive process. The main sources of carbohydrate in the human colon are dietary polysaccharides and glycosylated host secretions (Koropatkin, Cameron and Martens

2012). However, L reuteri is poorly equipped to degrade most of these carbohydrates (Gänzle and Follador 2012). The success of L. reuteri in the human gut might therefore depend on its ability to thrive on the remnants of primary fermenters, such as bifidobacteria, which are better equipped to utilize resources available in the distal intestinal milieu (Schell et al. 2002; Pokusaeva,

Fitzgerald and van Sinderen 2011; Sela et al. 2012).

Rhamnose forms part of the structural polysaccharides of the primary cell wall of plants (Yapo

2011). Fucose is abundant in human secretions as part of human milk oligosaccharides, mucins and other glycoconjugates in the intestinal epithelium (Becker and Lowe 2003; Zivkovic et al.

2011). Both of these deoxyhexoses reach the colon and are fermented into 1,2 PD by a number

135

of intestinal bacteria, making this compound readily available in the human gut (Saxena et al.

2010; Reichardt et al. 2014). Our results demonstrate that L. reuteri can obtain a growth advantage by crossfeeding from 1,2 PD produced by B. breve and E. coli from the fermentation of fucose and rhamnose, respectively. These results are in accordance with previous findings that show that in co-fermentation with glucose, 1,2 PD (and glycerol) increase the metabolic rate or L. reuteri resulting in higher cell yield (Talarico et al. 1990; Lin and Gänzle 2014; Rattanaprasert et al. 2014). By generating a knockout mutant we confirm that disproportionation of 1,2 PD into propanol and propionate is facilitated by the propanediol dehydratase PduCDE which is encoded in the human specific pdu-cluster (Morita et al. 2008; Frese et al. 2011). Analysis of the culture supernatants revealed that propanol is produced more abundantly than propionate (Table 3, Fig

3 and 5). This confirms that 1,2 PD serves as a hydrogen acceptor to recover NAD+, thus allowing

L. reuteri to spare the acetyl-phosphate to generate an extra ATP via production of acetate.( Table

3, Fig 6).

In the highly competitive ecosystem of the gastrointestinal tract, generating energy more efficiently can be key to the persistence of L. reuteri and might explain why human isolates have conserved the pdu-cluster (Frese et al. 2011). In this sense, it is important to point out that although L. reuteri has been regarded as autochthonous to the human gut (Reuter 2001), its prevalence appears to have reduced substantially in individuals of industrialized societies (Walter,

Britton and Roos 2011). Interestingly, a recent study found L. reuteri to be a dominant member of the fecal microbiome of natives of agriculturalist tribes consuming high fiber diets (Martínez et al.

2015). Research in mice has shown that insufficient supply of microbiota-accessible carbohydrates can irreversibly deplete species from the gut in just a few generations (Sonnenburg et al. 2016). The average modern diet is alarmingly low in fiber (Deehan and Walter 2016). A low intake of fiber would not only have consequences for primary fermenters reliant of these substrates for growth, but would also inevitably result in a decreased production of end products.

It is therefore possible that modern diets do not provide an adequate supply of fiber-derived

136

deoxyhexoses (i.e. rhamnose), resulting in a decreased production of 1,2 PD in the gut. This could mean that the niche conditions in which L. reuteri evolved with humans are no longer existent and would explain why human strains do not persist in the human, as shown in Chapter 2.

Taken together, findings generated in this study serve as proof-of-concept that a human-strain specific function might be relevant to the persistence to L. reuteri in the colon and provide novel insight into the molecular mechanisms that could underlie the association of L. reuteri with humans. Thus far, the relevance of pdu-cluster in human isolates of L. reuteri has been studied in the context of reuterin production and glycerol utilization (Lin and Gänzle 2014; Spinler et al.

2014). This is to our knowledge, the first demonstration of the relevance of the pdu-cluster for cross-feeding of 1,2D.

In vitro fermentations have proven to be excellent tools to study the metabolic underpinnings of cross-feeding amongst intestinal bacteria (Falony et al. 2006; Rakoff-Nahoum, Coyne and

Comstock 2014; Rivière et al. 2015; Turroni et al. 2015; Moens, Weckx and De Vuyst 2016;

Schwab et al. 2017). Nevertheless, these experiments are limited in that they do not replicate the intricacy of the intestinal ecosystem. Seminal findings regarding trophic interactions in the gut have been obtained using simplified in vivo models of cocolonization and in humanized mice

(Samuel and Gordon 2006; Mahowald et al. 2009). Future studies using similar approaches are thus warranted to determine of the relative importance of cross-feeding of 1,2 PD for L. reuteri in gut conditions.

137

Figure 4.6 Illustrative and schematic representation 1,2 PD cross-feeding and glucose metabolism in the presence of 1,2 PD (a) Illustrative representation of metabolic cross-feeding between 1,2 PD producers (E. coli and B breve) and L. reuteri. Dotted lines indicate cross-feeding of 1,2 PD. (b) General scheme of the heterofermentative metabolism of hexoses via de PKS pathway and 1,2 PD utilization by L. reuteri. The presence of 1,2 PD relieves the need for to utilize acetyl phosphate (acetyl-P) as a hydrogen acceptor to generate NAD+, thereby sparing this metabolite for ATP production. Major metabolic end products are printed in bold. Metabolic pathways drawn according to (Gänzle 2015)

4.5 Supplementary material

138

Table S.4.1 Oligonucleotides and plasmids used to generate the pduCDE mutant

Oligonucleotid Sequence Description es oVPL187 TACCGAGCTCGAATTCACTGG pVPL3002 Backbone, Rev oVPL188 ATCCTCTAGAGTCGACCTGC pVPL3002 Backbone, Fwd oVPL1335 GGTTCTTGAAATGTATGATGCA 5’ u/s flanking region of pduCDE, Fwd oVPL1336 ACAACAGCCGAACGATTTCC 5’ u/s flanking region of pduCDE, Rev oVPL1337 TGGTGCTAGCGAAATTGGAG 3’ d/s flanking region of pduCDE, Fwd oVPL1338 TACGATCTTGCCATTTTCAAC 3’ d/s flanking region of pduCDE, Rev oVPL1339 AGTGTAAAAGTTGAAAATGGCAAGATCGTAGG Bridging oligo for LCR TTCTTGAAATGTATGATGCATTGCGTCC oVPL1340 AAACGACGGCCAGTGAATTCGAGCTCGGTATG Bridging oligo for LCR GTGCTAGCGAAATTGGAGATACCATTGG oVPL1341 AGCGCTTATTGGAAATCGTTCGGCTGTTGTATC Bridging oligo for LCR CTCTAGAGTCGACCTGCAGGCATGCAA oVPL1342 AGTTGATGCCGGAGTACAAG u/s SCO screening, Fwd oVPL1343 TGGCGTGGCTTCATTGATTC u/s SCO screening, Rev oVPL1344 ACATTGGTTCCAGACTCACCAG d/s SCO screening, Fwd oVPL1345 ATGGCTGGACGTGAAGTAGG d/s SCO screening, Rev (sequencing) oVPL1346 TGAAGCCACGCCAGTAATTG (sequencing) oVPL1347 TGCAACGAAACCTTCTTCTGG (sequencing)

Plasmids pVPL3002 pORI19_EmR derivative with vancomycin counter-selection marker pVPL3478 pVPL3002 ::ΔpduCDE cassette

139

4.6 References

Arskold E, Lohmeier-Vogel E, Cao R et al. phosphoketolase pathway dominates in

Lactobacillus reuteri atcc 55730 containing dual pathways for glycolysis. J Bacteriol

2008;190:206–12.

Bäckhed F, Ley RE, Sonnenburg JL et al. Host-bacterial mutualism in the human intestine.

Science 2005;307:1915–20.

Baldomà L, Aguilar J. Metabolism of L-fucose and L-rhamnose in Escherichia coli: aerobic- anaerobic regulation of L-lactaldehyde dissimilation. J Bacteriol 1988;170:416–21.

Becker DJ, Lowe JB. Fucose: Biosynthesis and biological function in mammals. Glycobiology

2003;13, DOI: 10.1093/glycob/cwg054.

Belenguer A, Duncan SH, Calder AG et al. Two routes of metabolic cross-feeding between

Bifidobacterium adolescentis and butyrate-producing anaerobes from the human gut. Appl

Environ Microbiol 2006;72:3593–9.

Bunesova V, Lacroix C, Schwab C. Fucosyllactose and L-fucose utilization of infant

Bifidobacterium longum and Bifidobacterium kashiwanohense. BMC Microbiol 2016;16:248.

Clemente JC, Ursell LK, Parfrey LW et al. The impact of the gut microbiota on human health: an integrative view. Cell 2012;148:1258–70.

Coyte KZ, Schluter J, Foster KR. The ecology of the microbiome: Networks, competition, and stability. Science 2015;350:663–6.

Deehan EC, Walter J. The fiber gap and the disappearing gut microbiome: implications for human nutrition. Trends Endocrinol Metab 2016;27:239–42.

Duar RM, Frese SA, Fernando SC et al. Experimental determination of host adaptation of

Lactobacillus reuteri to different vertebrate species. Appl Environ Microbiol 2017:AEM.00132-17.

Duncan SH, Louis P, Flint HJ. Lactate-utilizing bacteria, isolated from human feces, that produce butyrate as a major fermentation product. Appl Environ Microbiol 2004;70:5810–7.

140

Egan M, O’Connell Motherway M, Kilcoyne M et al. Cross-feeding by Bifidobacterium breve

UCC2003 during co-cultivation with Bifidobacterium bifidum PRL2010 in a mucin-based medium.

BMC Microbiol 2014a;14:282.

Egan M, O’Connell Motherway M, Ventura M et al. Metabolism of sialic acid by

Bifidobacterium breve UCC2003. Appl Environ Microbiol 2014b;80:4414–26.

Falony G, Vlachou A, Verbrugghe K et al. Cross-feeding between Bifidobacterium longum

BB536 and acetate-converting, butyrate-producing colon bacteria during growth on oligofructose.

Appl Environ Microbiol 2006;72:7835–41.

Fischbach MA, Sonnenburg JL. Eating for two: how metabolism establishes interspecies interactions in the Gut. Cell Host Microbe 2011;10:336–47.

Flint HJ, Scott KP, Duncan SH et al. Microbial degradation of complex carbohydrates in the gut. Gut Microbes 2012;3:289–306.

Frese SA, Benson AK, Tannock GW et al. The evolution of host specialization in the vertebrate gut symbiont Lactobacillus reuteri. Genet 2011;7:16.

Frese SA, Mackenzie DA, Peterson DA et al. Molecular characterization of host-specific biofilm formation in a vertebrate gut symbiont. PLoS Genet 2013;9:e1004057.

Gänzle MG. Lactic metabolism revisited: metabolism of lactic acid bacteria in food fermentations and food spoilage. Curr Opin Food Sci 2015;2:106–17.

Gänzle MG, Follador R. Metabolism of oligosaccharides and starch in lactobacilli: a review.

Front Microbiol 2012;3:340.

Kok S De, Stanton LH, Slaby T et al. Rapid and reliable DNA assembly via ligase cycling reaction. ACS Synth Biol 2014;3:97–106.

Koropatkin NM, Cameron E a., Martens EC. How glycan metabolism shapes the human gut microbiota. Nat Rev Microbiol 2012;10:323–35.

Lin XB, Gänzle MG. Effect of lineage-specific metabolic traits of Lactobacillus reuteri on sourdough microbial ecology. Appl Environ Microbiol 2014;80:5782–9.

141

Mahowald MA, Rey FE, Seedorf H et al. Characterizing a model human gut microbiota composed of members of its two dominant bacterial phyla. Proc Natl Acad Sci U S A

2009;106:5859–64.

Martens EC, Lowe EC, Chiang H et al. Recognition and degradation of plant cell wall polysaccharides by two human gut symbionts. PLoS Biol 2011;9, DOI:

10.1371/journal.pbio.1001221.

Martínez I, Stegen JC, Maldonado-Gómez MX et al. The Gut Microbiota of Rural Papua New

Guineans: Composition, Diversity Patterns, and Ecological Processes. Cell Rep 2015;11:527–38.

Mazé A, O’Connell-Motherway M, Fitzgerald GF et al. Identification and characterization of a fructose phosphotransferase system in Bifidobacterium breve UCC2003. Appl Environ Microbiol

2007;73:545–53.

Moens F, Weckx S, De Vuyst L. Bifidobacterial inulin-type fructan degradation capacity determines cross-feeding interactions between bifidobacteria and Faecalibacterium prausnitzii.

Int J Food Microbiol 2016;231:76–85.

Morita H, Toh H, Fukuda S et al. Comparative genome analysis of Lactobacillus reuteri and

Lactobacillus fermentum reveal a genomic island for reuterin and cobalamin production. DNA Res

2008;15:151–61.

Oh JH, Van Pijkeren JP. CRISPR-Cas9-assisted recombineering in Lactobacillus reuteri.

Nucleic Acids Res 2014;42, DOI: 10.1093/nar/gku623.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Pokusaeva K, Fitzgerald GF, van Sinderen D. Carbohydrate metabolism in Bifidobacteria.

Genes Nutr 2011;6:285–306.

Rakoff-Nahoum S, Coyne MJ, Comstock LE. An ecological network of polysaccharide utilization among human intestinal symbionts. Curr Biol 2014;24:40–9.

142

Reichardt N, Duncan SH, Young P et al. Phylogenetic distribution of three pathways for propionate production within the human gut microbiota. ISME J 2014;8:1323–35.

Reuter G. The Lactobacillus and Bifidobacterium microflora of the human intestine: composition and succession. Curr Issues Intest Microbiol 2001;2:43–53.

Rivière A, Gagnon M, Weckx S et al. Mutual Cross-Feeding Interactions between

Bifidobacterium longum subsp. longum NCC2705 and Eubacterium rectale ATCC 33656 Explain the Bifidogenic and Butyrogenic Effects of Arabinoxylan Oligosaccharides. Schloss PD. Appl

Environ Microbiol 2015;81:7767–81.

Samuel BS, Gordon JI. A humanized gnotobiotic mouse model of host-archaeal-bacterial mutualism. Proc Natl Acad Sci 2006;103:10011–6.

Saxena RK, Anand P, Saran S et al. Microbial production and applications of 1,2-propanediol.

Indian J Microbiol 2010;50:2–11.

Schell MA, Karmirantzou M, Snel B et al. The genome sequence of Bifidobacterium longum reflects its adaptation to the human gastrointestinal tract. Proc Natl Acad Sci U S A

2002;99:14422–7.

Schwab C, Ruscheweyh H-J, Bunesova V et al. Trophic Interactions of Infant Bifidobacteria and Eubacterium hallii during L-Fucose and Fucosyllactose Degradation. Front Microbiol

2017;8:95.

Sela DA, Chapman J, Adeuya A et al. The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad

Sci U S A 2008;105:18964–9.

143

Sela DA, Garrido D, Lerno L et al. Bifidobacterium longum subsp. infantis ATCC 15697 - fucosidases are active on fucosylated human milk oligosaccharides. Appl Environ Microbiol

2012;78:795–803.

Seth EC, Taga ME. Nutrient cross-feeding in the microbial world. Front Microbiol 2014;5:350.

Sims IM, Frese SA, Walter J et al. Structure and functions of exopolysaccharide produced by gut commensal Lactobacillus reuteri 100-23. ISME J 2011;5:1115–24.

Sonnenburg ED, Smits SA, Tikhonov M et al. Diet-induced extinctions in the gut microbiota compound over generations. Nature 2016;529:212–5.

Spinler JK, Sontakke A, Hollister EB et al. From prediction to function using evolutionary genomics: human-specific ecotypes of Lactobacillus reuteri have diverse probiotic functions.

Genome Biol Evol 2014;6:1772–89.

Sriramulu DD, Liang M, Hernandez-Romero D et al. Lactobacillus reuteri DSM 20016 produces cobalamin-dependent diol dehydratase in metabolosomes and metabolizes 1,2- propanediol by disproportionation. J Bacteriol 2008;190:4559–67.

Stolz P, Böcker G, Hammes WP et al. Utilization of electron acceptors by lactobacilli isolated from sourdough. Zeitschrift fur Leb und -forsch 1995;201:91–6.

Talarico TL, Axelsson LT, Novotny J et al. Utilization of Glycerol as a Hydrogen Acceptor by

Lactobacillus reuteri: Purification of 1,3-Propanediol:NAD Oxidoreductase. Appl Environ Microbiol

1990;56:943–8.

Turroni F, Özcan E, Milani C et al. Glycan cross-feeding activities between bifidobacteria under in vitro conditions. Front Microbiol 2015;6:1030.

Walter J, Britton RA, Roos S. Host-microbial symbiosis in the vertebrate gastrointestinal tract and the Lactobacillus reuteri paradigm. Proc Natl Acad Sci U S A 2011;108 Suppl:4645–52.

Xu J, Bjursell MK, Himrod J et al. A Genomic View of the Human-Bacteroides thetaiotaomicron

Symbiosis. Science 2003;299:2047-6

144

Yapo BM. Rhamnogalacturonan-I: A structurally puzzling and functionally versatile polysaccharide from plant cell walls and mucilages. Polym Rev 2011;51:391–413.

Zivkovic AM, German JB, Lebrilla CB et al. Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci 2011;108:4653–8.

145

5. Chapter five: Identification and characterization of intestinal lactobacilli strains capable of degrading immunotoxic peptides present in gluten.

A version of Chapter five this chapter was published as: RM Duar, KJ Clark, PB Patil, C Hernández,

S Brüning, TE Burkey, N Madayiputhiya, SL Taylor, J Walter Journal of Applied Microbiology (2015) 118:

515-527

146

5.1 Introduction

Celiac disease is an increasingly diagnosed enteropathy caused by an uncontrolled immune response to wheat gluten and homologous proteins in barley and rye (Sollid 2002; Ludvigsson et al. 2013). The key T-cell stimulatory peptides responsible for celiac disease escape breakdown by gastrointestinal (GI) proteases and reach the small intestine where they cross the epithelial barrier to the lamina propria. The peptides become deaminated by the enzyme tissue transglutaminase and activate HLA-DQ2 and HLA-DQ8 restricted populations of CD4+ T cells, leading to inflammation and tissue damage (Green and Cellier 2007). Significant progress has been made in defining the epitopes involved in the pathogenesis of celiac disease (Arentz-Hansen et al. 2000; Shan et al. 2002; Shan et al. 2005; Tye-Din et al. 2010). Despite the absence of a single pathogenic motif, the gluten peptides involved in celiac disease are generally very rich in proline and glutamine residues, a feature that contributes to their resistance against proteolysis in the human gut (Shan et al. 2002).

The only available treatment for celiac patients is a strict life-long gluten-free diet. Compliance to a gluten-free diet represents a social and economic burden and the extensive use of gluten in processed foods, cross contamination, and improper labeling, hinders complete avoidance (Di

Sabatino and Corazza 2009). Recent studies have improved our understanding of the molecular basis of celiac disease and provided several targets for novel treatments (Sollid and Khosla 2005;

Kaukinen, Lindfors and Mäki 2014). One such approach might be the enzymatic detoxification of immunotoxic peptides during GI transit (Bethune and Khosla 2012). To date, enzymes evaluated for this purpose have been derived from plants, fungi, or bacteria that do not originate from GI environments (Piper, Gray and Khosla 2004; Shan et al. 2004; Siegel et al. 2006; Stepniak et al.

2006; Gass et al. 2007; Mitea et al. 2008). Although several of these enzymes efficiently detoxify gluten epitopes in vitro, their activity in the human GI tract remains to be assessed. To our knowledge, very few clinical trials have been conducted (Tack et al. 2013) and only one study has been successful (Lähdeaho et al. 2014)

147

The key to a complete removal of gluten toxic peptides in the small intestine might lie in the metabolic arsenal of the bacteria that naturally colonize the GI habitats. The mammalian GI tract harbors a complex, diverse and metabolically active bacterial community (Neish 2009). These microbes are capable of metabolizing complex dietary substrates non-digestible by the human host (Louis et al. 2007), and have been exposed to resistant gluten peptides for millennia. They are therefore likely to have evolved strategies to transport, internalize and utilize gluten peptides to satisfy their amino acid requirement (Davila et al. 2013). Supporting this hypothesis gluten degrading bacteria from the upper (Zamakhchari et al. 2011; Fernandez-Feo et al. 2013) and lower (Laparra and Sanz 2010; Caminero et al. 2014) human GI tract have been recently discovered. Moreover, imbalances in the composition and metabolism of the gut microbiota of children suffering from celiac disease have been reported, suggesting a potential role of gut bacteria in disease etiology (Tjellström et al. 2005; Nadal et al. 2007; Sanz et al. 2007). These findings clearly warrant further studies on both metabolic and immunologic aspects of gut bacteria in relation to celiac disease. Such bacteria could be used as probiotics with the goal to either remove epitopes before they reach the intestinal mucosa, to promote epithelial healing, or to directly target pathological immune responses (Lindfors et al. 2008; Laparra and Sanz 2010; de

Sousa Moraes et al. 2014)

The goal of this study was to identify and characterize bacterial isolates from the proximal GI tract of pigs capable of degrading peptides involved in the etiology of celiac disease. Pigs were used in our study as these animals, in contrast to humans, harbor high numbers of lactic acid bacteria (LAB), especially lactobacilli and streptococci, in their stomach and small intestine (up to

109 bacteria/gram content) (Castillo et al. 2007). These bacteria are adapted to the conditions in the proximal GI tract and might be good candidates for probiotics aimed at removing gluten epitopes before they reach the epithelium of the small intestine in celiac patients. We successfully identified four strains of lactobacilli that showed degradation of several well characterized gluten immunotoxic peptides (Shan et al. 2002; Tye-Din et al. 2010). We tested the peptide degradation

148

activity of these strains alone or in combination and we also compared it with strains commonly used as probiotics.

5.2 Materials and Methods

5.2.1 Animals and intestinal sampling

Housing, and management of animals was according to guidelines set forth by the University of Nebraska-Lincoln Institutional Animal Care and Use Committee and described in animal care protocol #07-04-014D.

Weaned pigs (n = 6, initial BW = 4.5 - 7.0 kg; 10 weeks of age) were fed a basal corn-soy diet supplemented with 20% gluten (Vital Wheat Gluten, ADM Milling Co) for at least 16 weeks to stimulate gluten-utilizing bacterial populations in the GI tract. Animals were sacrificed at 26 weeks at the UNL Loeffel Meat Laboratory (a USDA-inspected facility) according to approved procedures. Immediately after sacrifice, 100 ml of proximal small intestinal (SI) contents were collected from each animal. Large debris was removed from the SI contents by low speed centrifugation (200 x g for 5 min). Samples were then centrifuged at 6000 x g for 20 min and pellets resuspended in 5 ml PBS. Aliquots of 1.5 ml were mixed with 0.5 ml of 60% glycerol and stored at -80 °C.

5.2.2 Enrichment of gluten utilizing bacteria

To enrich for strains capable of metabolizing gluten peptides, a gluten peptide enrichment

(GPE) media was devised. This medium was a chemically defined minimal media as described by Elli et al. (2000) with some modifications. The amino acids glutamine, leucine, proline, phenylalanine, and tyrosine were replaced by 5 g/l of the 18-mer GLIAα-2/9 peptide

LQLQPFPQPQLPYPQPQL (Arentz-Hansen et al. 2000), which is also part of the 33-mer (Shan et al. 2002). The GPE media contained glucose 10.0 g/l, yeast extract 0.5 g/l, potassium hydrogen phosphate 3.1 g/l, di-ammonium hydrogen citrate 2.0 g/l, potassium di-hydrogen phosphate 1.5

149

g/l, sodium chloride 0.02 g/l1, ascorbic acid g/l, potassium acetate 10.0 g/l, tween 80 1.0 g/l, heptahydrated magnesium sulphate 0.5 g/l, hydrated manganese sulphate 0.02 g/l , cobalt sulphate 0.01 g/l, calcium lactate 1.0 g/l, DL-2-aminobutyric acid 0.1 g/l, L-serine 0.1 g/l, L- threonine 0.1 g/l, L-cysteine 0.1 g/l, L-asparagine 0.1 g/l, L-isoleucine 0.1 g/l, L-methionine 0.1 g/l, L-tryptophan 0.1 g/l, L-valine 0.1 g/l , DL-alanine 0.2 g/l, L-aspartic acid 0.3 g/l, glycine 0.2 g/l,

L-histidine-HCL 0.2 g/l, L-lysine-HCL 0.2 g/l, L-arginine 0.2 g/l, guanine 0.1 g/l, thymine 0.1 g/l,

cytidine 0.1 g/l, 2'-deoxyadenosine 0.1 g/l and 2'-deoxyuridine 0.1 g/l. FeSO4 was added at the concentration of 0.02 g/l immediately before using the media. The pH was adjusted to 6.9 to resemble small intestine conditions. Frozen stocks were pooled, centrifuged (6000 x g, 5 min), and supernatants removed. Pellets were resuspended in 1.0 ml of PBS and used to inoculate the

GPE medium at 0.8%. Enrichment cultures were incubated anaerobically at 37 ºC for 24 hours, before a second enrichment step was performed with 1% inoculum under the same conditions.

Bacterial population dynamics during the enrichment process with and without the addition of the peptide were characterized by denaturing gradient gel electrophoresis (DGGE) as previously described (Martínez et al. 2009).

5.2.3 Culture techniques and classification of isolates

Bacteria were cultured from fresh SI contents and from the 24 hour and 48 hour GPE, by dilution plating on Rogosa (selective for lactobacilli), BEA (enterococci), M17 (non-selective, isolation of streptococci), and BHI (universal) agar. Colonies were picked and grown at least once in liquid media before dilution streaking to assure purity for preparation of stock cultures.

Chromosomal DNA from pure cultures was extracted using the DNeasy® Blood and Tissue kit

(Qiagen) with modifications according to Oh et al. (2010). Isolates were typed by RAPD-PCR as described by Meroth et al. (2003). Cultures with unique RAPD patterns were taxonomically classified by 16s RNA gene sequencing and database comparisons using the Ribosomal

150

Database Project SEQ MATCH tool (http://rdp.cme.msu.edu/seqmatch/seqmatch_intro.jsp) and the NCBI nucleotide Blast web tool (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

5.2.4 Initial screen of isolates for specific proteolytic activities

The chromogenic substrates Suc-Ala-Pro-p-NA, Leu-p-NA and Pro-p-NA (Sigma Chemical

Co. St Lois, MO, USA) were used to screen isolates for the production of prolyl endopeptidase

(PEP), aminopeptidase type N (PepN), and proline iminopeptidase (PepI), respectively.

Proteolytic activities were tested as described previously (Rollán et al. 2005), with some modifications. Bacterial strains were grown in their respective media and sub-cultured twice.

Twelve hours cells were harvested by centrifugation (7000 x g for 7 min at 4 °C) and washed twice with 0.05 mol/l tris buffer (pH 7.0). Cells were re-suspended in 1.0 ml of lysis buffer (sucrose

24%, lysozyme 10.0 mg/ml and mutanolysin 50 U/ml) and incubated at 37 °C for 30 min. Cell walls were disrupted using bead beating for 2 min followed by centrifugation (7000 x g for 10 min at 4 °C) to obtain cell-free extracts. Protein concentration was determined using the Bradford test and standardized to 2.5 mg/ml. The enzyme reaction mixture contained 360 µl of 0.02 mol/l

-1 phosphate buffer (pH 7.0), 150 µl of substrate (0.007 mol l ), 8 µl of NaN3 (0.05% final concentration), and 50 µl of cytoplasmic extract. Samples were incubated at 37 °C. Release of the p-nitroaniline molecule was measured spectrophotometrically at 410 nm at zero and after four hours.

5.2.5 Immunotoxic peptides

Peptides with a demonstrated importance in the etiology of celiac disease were synthesized by Bio-Synthesis Inc. (Texas, USA). Included in this study were α-gliadin derived peptides 33-mer

LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF (Shan et al. 2002), and 16-mer

QLQPFPQPQLPYPQPQ (Tye-Din et al. 2010), and the ω-gliadin/C-hordein derived 17-mer

QPQQPFPQPQQPFPWQP (Tye-Din et al. 2010).

151

5.2.6 Bacterial strains, culture media, and growth conditions

In addition to the bacteria isolated from the pig SI contents and the GPE enrichment during this study, the following commercially available probiotic strains were included: Lactobacillus plantarum 299v, L. casei ATCC 334, L. acidophilus NCFM, Bifidobacterium animalis BB-12, L. reuteri ATCC 55730, L. rhamnosus GG and the probiotic preparation VSL#3 containing, L plantarum, L. acidophilus, L. casei, L. delbrueckii spp. bulgaricus, Streptococcus thermophilus, B. breve, B. longum and B. infantis. Bifidobacteria and lactobacilli strains were cultured anaerobically on MRS and modified MRS media (MRS supplemented with 10.0 g/l maltose and 5.0 g/l fructose), respectively. Streptococcus thermophilus was grown on M17 media under aerobic conditions. All incubations were conducted at 37°C.

5.2.7 Peptide degradation

Peptide degradation was determined in vitro under conditions reflective of the small intestine.

In brief, one colony from a 48 hour agar plate was transferred into broth media and incubated overnight. Fresh media was inoculated with 1% of the overnight culture and incubated for 16 hours. Cells were collected by centrifugation and individual isolates or an even mixture of the strains were re-suspended to an OD620 of 2.0 in minimal media (Elli et al. 2000) at pH 7.0 containing glucose 10.0 g/l, potassium hydrogen phosphate 3.1 g/l, di-ammonium hydrogen phosphate 2.0 g/l, potassium dihydrogen phosphate 1.5 g/l, sodium chloride 0.02 g/l, ascorbic acid 0.5 g/l, potassium acetate 10.0 g/l, calcium lactate 1.0 g/l, and 1.0 mg/ml of one immunotoxic peptide as the sole amino acid source. Samples were incubated anaerobically with constant shaking (150 rpm) at 37 °C. After 24 hours, bacterial cells were removed by centrifugation (10000 x g for 10 min at 4 °C) and trifluoroacetic acid was added to a final concentration of 0.1% (v/v) to stop enzymatic reactions. Control samples without bacterial cells were prepared in parallel and

152

collected at zero and after 24 hours under the same conditions as test samples. All samples were stored at -20 °C until analyzed.

5.2.8 ELISA tests

The concentration of all three peptides was determined with the RIDASCREEN® (R-Biopharm

AG, Darmstadt, Germany) competitive ELISA kit. The peptides 33-mer and 16 mer were also quantified with the GlutenTox® (Biomedal, Seville, Spain) ELISA kit. Samples were analyzed according to the manufacturer’s instructions, but quantification was performed using standard curves generated with the respective peptides (Tables S1 – S2). Degradation was calculated as the percent decrease in peptide concentration (µg/ml) in the 24 h test sample relative to the 24 h control sample (without bacterial inoculum).

5.2.9 Mass Spectrometry

Peptide degradation during the GPE enrichment and during the initial screening process was measured by ultrahigh performance liquid chromatography and tandem mass spectrometry in multiple reaction monitoring mode (UHPLC-MS/MS-MRM) at the Proteomics and Metabolomics

Core Facility, Department of Biochemistry (University of Nebraska, Lincoln, USA). The UHPLC

(Agilent Technologies, Palo Alto, CA) was coupled to a triple quadrupole mass spectrometer (Q-

Trap 4000, AB SCIEX) and operated in positive ion ESI-MRM more. Gradient reverse phase chromatography separation of the peptide was carried out at a flow rate of 300 µl min-1. Peptides were eluted on 0.1% formic acid (buffer A) over a 15 min interval using a mixed gradient (0 min =

1%, 7 min = 95%, 10 min = 1%) of buffer B containing 0.1% formic in acetonitrile. Injection volume was 10 µl. Analyst 1.4.2 software was used for instrumentation control and for MRM quantitative analysis based on calibration curves.

Amino acid sequences of the fragments originating from peptide hydrolysis by the four strains selected during this study were determined at the Mass Spectrometry Facility at the National

153

Jewish Medical and Research Center (Denver, CO, USA) using a quadrupole time of flight mass spectrometer (Agilent Technologies. Palo Alto, CA) equipped with an HPLC-Chip Cube system operated in mass spectrometry (LC-MS) and tandem mass spectrometry (LC-MS/MS) modes. In brief, aliquots (10 µl) from the peptide test samples and the 24 hour control were vacuum dried at

45 ºC and stored at -20 ºC. Samples were resuspended in 100 µl of 3% acetonitrile + 0.1% formic acid in water. Injection volumes were 0.5 µl and 2.0 µl for the LC-MS and LC-MS/MS analysis, respectively. Peptides were separated with a gradient (0 min = 3%, 0.5 min = 10%, 7 min = 50%,

7.1 min = 80%, 9 min = 80%, 10 min = 3%) of buffer A, containing 0.1% formic acid, and buffer B containing 0.1% formic acid in 90% acetonitrile, at a flow rate of 0.45 µl/min. Mass, abundance and retention time were obtained from the LC-MS analysis using the Agilent Technologies Mass

Hunter Qualitative Analysis software, version B.06.00. Fragment sequences were obtained by searching an in-house library using the Spectrum Mill software (Rev A.03.03.038 SR1). Cleavage sites were determined based on fragment sequence and abundance (peak area) (Tables S5 –

S7).

5.2.10 Statistical analysis

Initial screening for specialized proteolytic enzymes, ELISA tests and UHPLC-MS/MS-MRM analysis were carried out in duplicate. RIDASCREEN® ELISA data from the selected strains and the commercial probiotics were analyzed in triplicate. All data are shown as the average ± standard deviation. Peptide fragment abundance was calculated as the average peak area from the LC-MS/MS analysis of two independent samples. Correlations between peptide peak area

(UHPLC-MS/MS-MRM) and staining intensity of the DGGE fragments were assessed by

Pearson's correlation test (α = 0.05) using GraphPad Prism version 6.04 (GraphPad Software,

San Diego, CA).

154

5.3 Results

5.3.1 Isolation of intestinal bacteria directly from the proximal GI tract of pigs and after enrichment

We employed in vivo and in vitro enrichment procedures with the goal to isolate bacteria capable of metabolizing gluten peptides (Fig. 5.1). Pigs were fed a basal diet supplemented with

20% gluten for 16 weeks to enrich for intestinal bacteria in vivo. In addition, an enrichment step was performed in a defined growth medium (GPE) containing the toxic gluten peptide

LQLQPFPQPQLPYPQPQL, but not the individual amino acids included in the peptide. In total,

140 isolates were obtained, 87 directly from the SI contents and 53 isolates from the enrichment

(Fig. 5.1). All isolates were typed by RAPD and 96 isolates with distinct patterns were taxonomically identified by 16S rRNA gene sequence analysis.

In accordance with the microbiota composition of the proximal gut of pigs (Leser et al. 2002), the strains isolated directly from the SI contents, belonged to the genera Lactobacillus (33%),

Clostridium (20%), Escherichia (18%) and Streptococcus (13%) with the remaining belonging to

Bacillus (10%) and other genera (6%). After the GPE step, lactobacilli accounted for 57% of the strains isolated, with the majority of the isolates belonging to the species L. ruminis and L. amylovorus. Other genera included Enterococcus (40%) and Streptococcus (3%). Therefore, it appears that the in vitro enrichment favored the growth of LAB, particularly lactobacilli and enterococci. Comparison of the bacterial populations in the GPE media with and without the peptide by DGGE confirmed the enrichment of several bacterial species, including Lactobacillus and Enterococcus, through the 18-mer (Fig. 5.2a). Quantification via UHPLC-MS/MS-MRM also revealed that the peptide was gradually degraded during the enrichment step (Fig.5.2b).

Interestingly, peptide concentration showed a significant negative correlation (p < 0.005) with the staining intensity of a DGGE fragment corresponding to L. amylovorus (Fig. 5.2c).

155

Figure 5.1 Experimental design

In vivo (in pigs) and in vitro enrichment procedures used in this study to select gluten- degrading bacteria adapted to the proximal GI tract. Number of strains selected from each step is shown in brackets.

156

5.3.2 Screening for isolates with peptidase activity

All 96 unique strains based on RAPD analysis were screened for the production of specialized peptidases known to degrade immunotoxic gluten peptides (Piper, Gray and Khosla 2004; De

Angelis et al. 2010). PepN and PEP were the main enzymatic activities observed among the strains, while PepI activity was lower but also detected in a number of isolates (Fig. 5.S1). Based on this analysis, 12 isolates with the highest overall activity were selected for in depth characterization.

5.3.3 Degradation of peptides involved in the etiology of celiac disease

To evaluate the potential of the selected 12 strains to detoxify gluten epitopes, we determined the degradation of the metastable 33-mer, 16-mer, and 17-mer peptides, which have been shown to be involved in the etiology of celiac disease (Shan et al. 2002; Tye-Din et al. 2010). Two different quantitative ELISA assays, in addition to UHPLC-MS/MS-MRM analysis, were used to quantify the peptides before and after 24 hour incubation with the bacterial strains. This analysis revealed that several of the strains degraded the three peptides, but differed markedly in their hydrolytic activity towards the immunogenic peptides (Fig.5.S2-S4). Based on the combined results from these assays, four gluten peptide degrader (GPD) strains (L. amylovorus GPD2, L. johnsonii GPD6, L. ruminis GPD9, and L. salivarius GPD12) with the highest rates of degradation were selected.

The selected GPD isolates were characterized in depth for their ability to hydrolyze the gluten peptides as individual strains and as a mixture (Fig. 5.3). The RIDASCREEN® ELISA was used to measure degradation, as the R5 antibody detected all gluten peptides studied (Table S3). As shown in Figure 5.3, the findings were in general agreement with those from the initial screen, and all GPD isolates hydrolyzed the gluten peptides at different rates. Specifically, the hydrolytic actions of the strains L. salivarius GPD12 and L. amylovorus GPD2 were the most effective in

157

degrading the 33-mer and the 17-mer peptides, while the 16-mer was more effectively degraded by the strains L johnsonii GPD6 and L. ruminis GPD9. The GPD strains were mixed to test whether the diverse proteolytic specificity shown could be combined to achieve a synergistic effect. The amount of degradation achieved was typical of the strain that best degraded that particular peptide. However, the mixture of strains produced an overall more uniform degradation of all three peptides than the individual strains GPD2, GPD6 and GPD9, and outperformed the strain GPD12 in degradation of the 16-mer.

Figure 5.2 Microbiological and analytical characterization of the bacterial enrichment cultures (a) DGGE during incubation without (left) and with (right) the 18-mer peptide. Corresponding bands are indicated by colored circles. (b) UHPLC-MS/MS-MRM analysis of the first 24 hours of enrichment. Peak corresponding to the 18-mer is indicated with a blue box. (c) Correlation between peak area and the DGGE band intensity of L. amylovorus.

158

Figure 5.3 Degradation of the gluten peptides by the selected gluten peptide degrader GPD strains 33-mer (black), 17-mer (grey), 16-mer (white). Bars represent the average percent degradation (n=3 ± SD) after 24 hours determined from peptide amounts in test and control samples quantified using the RIDASCREEN® ELISA.

159

5.3.4 Characterization of cleavage fragments and proteolytic specificity

To evaluate to which degree the GPD strains removed T-cell stimulating epitopes from peptides, the amino acid sequences of the bacterial hydrolysis fragments were determined.

According to the LC-MS/MS results, all the GPD strains partially hydrolyzed the highly resistant

33-mer mainly by targeting post proline and post glutamine bonds (Table S5). But more importantly, at least one cleavage site was found in all of the six overlapping epitopes producing a variety of peptide fragments with a reduced number of immunodominant T-cell epitopes (Table

6.1 and Fig. 5.4). As shown in Figure 5.4, hydrolysis of the 33-mer peptide was strain-dependent and cleavage patterns were in general accordance with enzyme production levels by each GPD strain. Specifically, the PepN from L. amylovorus GPD2 (isolate 50 in Fig. 5.S1) targeted the bonds at the N-terminal of the peptide, generating fragments of 8 to 28 amino acids (Fig. 5.4b).

Three short fragments, lacking 9 amino acid T-cell stimulatory core epitopes, were also produced.

Hydrolytic actions of L. johnsonii GPD6 (isolate 19 in Fig. 5.S1) reflected the endoproteolytic activity of PEP in combination with PepN cleavage, producing medium sized fragments with one to 5 epitopes, and several small fragments (8 – 13 aa) (Fig. 5.4c). L. ruminis GPD9 (isolate 44 in

Fig 5.S1), a high producer of PepN and PEP, cleaved at both ends of the 33-mer peptide, leaving a few short peptide fragments absent of known T-cell stimulating epitopes, and several other fragments containing as many as five epitopes (Fig. 5.4d). Hydrolysis by the strain L. salivarius

GPD12 (isolate 59 in Fig 5.S1) produced the highest reduction of the native form of the 33-mer generating the truncated peptide PQPQLPYPQPQLPYPQPQLPYPQPQP. Reflective of PepN activity, hydrolysis occurred at the N-terminus of the peptide. Other degradation fragments ranged from 7 to 28 amino acids, the majority containing two or more epitopes (Fig. 5.4d). The hydrolytic activity of the GPD mixture produced truncated 33-mer peptide fragments ranging from14 to 26 aa. Smaller fragments (< 8 aa) were not detected (Fig. 5.4e).

160

Figure 5.4 Characterization of peptide degradation by LC-MS/MS (a) Peptide sequence of the 33-mer with immunostimulatory T cell epitopes underlined (Sollid et al. 2012). Peptide fragments after hydrolysis by (b) L. amylovorus GPD2, (c) L. johnsonii GPD6, (d) L. ruminis GPD9, (e) L. salivarius GPD12, and (f) the GPD mix. Relative fragment abundance is represented by a colored scale and the remaining T-cell epitopes are indicated in colored boxes same as (a).

161

Hydrolysis of the 17-mer peptide was less proline specific and primarily targeted the bond after the Q5 residue, which is located outside the 9 amino acid epitope core. Fragments generated ranged from 11 to 14 amino acids (Table S6) and only L. salivarius GPD12 cleaved the bonds

P↓F and P↓W located within the two overlaping epitopes (Table 6.1).The 16-mer was hydrolyzed into smaller fragments of 9 or 10 residues. The most common cleavage site was between the F5 and the P6 residues generating the fragment PQPQLPYPQPQ. Hydrolysis with L. amylovorus

GPD2 and L. salivarius GPD12 also generated the fragment QPQLPYPQPQ (Table S7). All the

GPD strains with the exception of L. johnsonii GPD6, which cleaved at only one of the epitopes, hydrolyzed bonds contained within the two overlapping epitopes of the 16-mer peptide (Table

6.1). In general, peptide degradation by the GPD mix reflected the hydrolysis produced by the individual strains with the exception of the cleavage site between Q29 and P30 of the 33-mer which was not produced by any of the GPD isolates alone (Table 6.1).

Table 5.1 Cleavage sites in the immunotoxic peptides by bacterial strain or strain mixture 33mer* 17mer* 16mer* LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF QPQQPFPQPQQPFPWQP QLQPFPQPQLPYPQPQ

L. amylovorus LQLQPFPQPQLPYPQPQLPYPQPQLPYPQP QPQQPFPQPQQPFP QLQPFPQPQLPY GPD2 QPF WQP PQPQ L. johnsonii LQLQPFPQPQLPYPQPQLPYPQPQLP QPQQPFPQPQQPF QLQPFPQPQLPYPQ GPD6 YPQPQPF PWQP PQ

L. ruminis LQLQPFPQPQLPYPQPQLPYPQPQL QPQQPFPQPQQPFP QLQPFPQPQLPYPQ GPD9 PYPQPQPF WQP PQ

L. salivarius LQLQPFPQPQLPYPQPQLPYPQPQ QPQQPFPQPQQP QLQPFPQPQLP GPD12 LPYPQPQPF FPWQP YPQPQ

GPD mix† LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQ QPQQPFPQPQQP QLQPFPQPQLPY PF FPWQP PQPQ

*T-cell stimulatory epitopes (Tye-Din et al. 2010; Sollid et al. 2012) are indicated with a black line.  Arrows indicate major cleavage sites Arrows indicate less efficiently cleaved sites. † Combination of all four isolates

162

5.3.5 Comparison with commercial probiotic strains

Peptide degradation abilities of strains obtained during this study were compared with those of bacteria commercially used as probiotics using the RIDASCREEN® ELISA as described above.

In general, the probiotic strains differed widely in their ability to degrade the peptides and some

(B. animalis BB-12, L. reuteri ATCC 55730) showed very little or no hydrolytic activity. L. rhamosus

GG (all peptides), L. casei ATCC 334 (33-mer and 16-mer), L. plantarum 299v (33-mer and 17- mer) and L. acidophilus NCFM (16-mer), showed similar degradation to the GPD strains (Table

S4). However, the mixture of GPD isolates outperformed the eight members of the probiotic preparation VLS#3 in the hydrolysis of the immunogenic peptides (Table S4).

5.4 Discussion

The objective of this study was to characterize the ability of bacteria from the proximal GI tract of pigs, to degrade toxic gluten epitopes in order to identify candidates for probiotic applications.

Our approach involved in vivo and in vitro enrichment procedures to select for gluten degrading bacteria. Pigs are physiologically similar to humans (Heinritz, Mosenthin and Weiss 2013) but harbor higher numbers of LAB in the proximal GI tract (Castillo et al. 2007). These animals are an excellent model for humans and their use allowed to feed high amounts of gluten to stimulate gluten-metabolizing bacteria in the small intestine, a region that harbors only very low numbers of bacteria in humans (Walter 2008). Analytical and microbiological characterization of the bacterial population dynamics during and after the in vitro enrichment, suggest lactobacilli had a competitive advantage over other intestinal bacteria, conferred by their ability to utilize metastable gluten peptides to obtain nitrogen.

Our extensive screening approach further confirmed this hypothesis as the strains with the highest gluten-degrading abilities belonged to the species L. amylovorus, L. johnsonii, L. ruminis, and L. salivarius. These strains degraded the metastable peptides 33-mer, 16-mer, and 17-mer, which have been shown to be relevant for the immune response in celiac disease (Shan et al.

163

2002; Tye-Din et al. 2010). Our findings are in accordance to those of Caminero et al. (2014) in that we also identified lactobacilli as the dominant gluten degrading bacteria in the GI tract.

Interestingly, both studies have identified L. ruminis and L. amylovorus as the major gluten degraders, supporting the role of these species in gluten degradation.

Particular aspects of the physiology of lactobacilli could assist in the degradation of gluten.

LAB produce various peptidases aimed to obtain amino acids from the nutrient-rich environments where they typically inhabit (Fernández and Zúñiga 2006; Makarova et al. 2006). Their limited biosynthetic metabolism is compensated by an array of hydrolytic enzymes and transport systems

(Makarova et al. 2006). To satisfy their nitrogen needs, extracellular and cell-wall bound proteases cleave proteins into oligopeptides that are transported across the membrane to be hydrolyzed by intracellular peptidases into free amino acids (Christensen et al. 1999). The proteolytic enzymes expressed by lactobacilli are capable of cleaving proline bonds, which was evidenced by our findings and their ability to utilize the proline-rich protein casein (Christensen et al. 1999). In fact, various strains of LAB degrade toxic gluten peptides during sourdough fermentation (De Angelis et al. 2006, 2010; Gobbetti et al. 2007; Rizzello et al. 2007).

The degree by which the GPD isolates degrade immunogenic peptides appears to be peptide- dependent, and it is likely to involve both extracellular proteolysis and peptide uptake. Specifically, hydrolysis of the 33-mer was the lowest among the three peptides when analyzed by ELISA.

However, LS-MS/MS analysis revealed the production of several medium-sized and short fragments that were detected in low abundance. On the contrary, significant reduction of the 16- mer and 17-mer peptides was detected by ELISA, but fewer and only slightly truncated fragments were produced. This trend was observed in all the GPD isolates. Therefore, it is possible that the longer 33-mer peptide was target of extracellular hydrolysis and that intact or slightly truncated

17-mer and 16-mer peptides were transported across the membrane to be cleaved in the cytoplasm. It has been well established that some LAB are capable of transporting proline rich oligopeptides of various lengths through the cell membrane (Berntsson et al. 2009) and several

164

species of lactobacilli, encode for these transporters in their genomes (Liu et al. 2010). However, more functional studies are required to determine how the GPD strains remove the metastable peptides.

It is important to point out that, although the GPD strains can hydrolyze metastable gluten peptides, the degradation products are likely still immunogenic. However, partial cleavage, especially at the proline level could potentially reduce peptide resistance to hydrolysis by digestive enzymes, and it might constitute a beneficial trait if probiotic strains are used in combination with gluten-degrading enzymes in clinical applications (Tye-Din et al. 2010; Siegel et al. 2012; Tack et al. 2013; Lähdeaho et al. 2014). Lactobacillus strains might have synergistic activities with enzymes to completely destroy the immunogenic epitopes by complementing their cleavage range. For example, the GPD stains mainly cleaved at the N and C terminus of the peptides leaving immunogenic internal fragments intact. Combination of the GPD strains with enzymes that preferably cleave at internal bonds (Stepniak et al. 2006) could be act synergistically to completely destroy peptide epitopes. Additionally, fragments generated from partial peptide hydrolysis could be subsequently cleaved by enzymes that show preference for short immunogenic substrates (Shan et al. 2004). Further research is required to determine if combinations of enzymes and GPD bacteria successfully eliminate peptide immunogenicity. The information presented in this study can hereby serve as a guide for the selection of strains to be combined with specific gluten-degrading probiotics and enzymes.

This work, together with recently published reports (Laparra and Sanz 2010; Zamakhchari et al. 2011; Fernandez-Feo et al. 2013; Caminero et al. 2014) provides a basis for the selection of probiotic strains as a potential treatment for celiac patients (de Sousa Moraes et al. 2014).

Delivery of the probiotic strains could be achieved in the form of capsules administered before and after a meal to assure the presence of lactobacilli at the time of gluten exposure. Probiotic lactobacilli (including strains from pigs) persist in the human gut for a number of days (Jacobsen et al. 1999) and several LAB strains haven been successfully used as delivery vehicles of mucosal

165

therapeutic biomolecules (Daniel et al. 2011). These findings are encouraging in that they suggest that lactobacilli could be delivered, in a functional state, to the relevant locations in the gut. Given the incomplete degradation of the immunotoxic peptides by the GPD strains identified in our study, the use of a probiotic treatment alone is unlikely to allow patients to consume significant amounts of gluten. However, probiotics could potentially aid to control the detrimental effects from accidental exposure and contamination of bona fide gluten-free products, and they could be combined with gluten degrading enzymes with synergistic activity. However, research is needed to determine the gluten degrading potential lactobacilli in vivo, and to develop therapeutic formulations and dosing regimens for clinical applications. Probiotic lactobacilli could be used in human clinical research without extensive hurdles, as these bacteria are generally harmless and well tolerated (even if consumed in high numbers), and various strains have a “generally recognized as safe” status (GRAS) (Wells and Mercenier 2008). In this study, we showed that several commercially available probiotic strains (L. rhamnosus GG, L. casei 334, and L. plantarum

299v), that could be readily applied, also degrade the immunotoxic peptides to some extent.

However, the four porcine stains identified during our strain selection might have advantages to human colonic (Laparra and Sanz 2010; Caminero et al. 2014), oral (Zamakhchari et al. 2011;

Fernandez-Feo et al. 2013), and genetically engineered bacteria (Alvarez-Sieiro et al. 2014) in that they are naturally adapted to the conditions of the small intestine, and therefore are likely more physiologically active and functional. However, ultimately, only clinical trials can determine which strains work best in humans.

166

5.5 Supplementary material

Table S2 Sample dilution and standard curve for RIDASCREEN® ELISA Peptide Sample Std. 1 Std. 2 Std. 3 Std. 4 Std. 5 dilution (µg ml - (µg ml - (µg ml - (µg ml - (µg ml 1 ) 1 ) 1 ) 1 ) -1) 17-mer 1:100 10 3.3 1.1 0.4 0.1 16-mer 1:7 143 47.7 15.9 5.3 1.7 33-mer 1:5 200 66.7 22.2 7.4 2.5

Dilutions were prepared in the sample dilution buffer provided with the kit

Table S3 Sample dilution and standard curve for the GlutenTox® ELISA Peptide Sample Std. 1 Std. 2 Std. 3 Std. 4 Std. 5 dilution (µg ml - (µg ml - (µg ml - (µg ml - (µg ml 1 ) 1 ) 1 ) 1 ) -1 ) 16-mer 1:5000 200 100 50 25 12.5 33-mer 1:5000 200 100 50 25 12.5

Dilutions were prepared in the sample dilution buffer provided with the kit

Table S3 Epitopes recognized by the ELISA tests.

Peptide Amino acid sequence 33 mer. LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 17 mer. QPQQPFPQPQQPFPWQP 16 mer. QLQPFPQPQLPYPQPQ Epitopes recognized by the monoclonal antibodies R5 (RIDASCREEN®) and G12 (GlutenTox®) are shown in bold or italicized, respectively

Table S4 Percent peptide degradation by the GPD and commercial probiotic strains L. B. L. L. L. L. L. L. L. saliva anim amylov johnso rumini planta casei acidop reuteri L.rhamn GPD rius alis VSl#3 orus nii s rum ATCC hilus ATCC osus GG mix GDP1 BB- GPD2 GPD6 GDP9 299v 334 NCFM 55730 2 12 33- 28.0 ± 5.0 ± 40.0 ± 41.0 ± 19.0 ± 5.3 ± 2.7 ± 30.0 ± 4.5 ± 8 ± 2.6 0 ± 0 26 ± 21 mer 3.5 1.0 4.0 23 7.5 9.2 4.6 2.1 2.6 17- 44.0 ± 25 ± 19 ± 52 ± 37 ± 0.67 ± 0.67 ± 1.7 ± 0.67 ± 53.0 ± 43.5 ± 35.0 ± mer 2.1 1.7 3.1 2.3 1.2 0.58 1.2 2.9 0.58 13.0 5.5 10.2 16- 24.0 ± 66.0 ± 56.0 ± 35.0 ± 0.67 ± 1.7 ± 0.67 ± 65.0 ± 19 ± 41 ± 18 48 ± 37 25 ± 6.4 mer 2.1 16.0 29.0 3.6 0.58 1.2 0.58 17.1 4.2

167

168

Fig. S1 Specialized peptidase production measured as the release of p-nitroaniline from chromogenic substrates. Bars represent absorbance at 410 nm after 4 hours. Prolyl endopeptidase (black), aminopeptidase N (grey), and proline iminopeptidase (white). Isolates selected for gluten peptide degradation experiments are indicated with an arrow ()

169

Table S5. MS-MS data from 33-mer hydrolysis by each strain start aa sequence area 1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 1.67E+08 1 LQLQPFPQPQLPYPQ 2.68E+06 6 FPQPQLPYPQPQLPYPQPQLPYPQPQPF 4.03E+05 7 PQPQLPYPQPQLPYPQPQLPYPQPQPF 9.86E+06 12 PYPQPQLPYPQPQLPYPQPQPF 2.77E+06 L amylvorus 15 QPQLPYPQPQLPYPQPQPF 1.02E+06 GDP2 17 QLPYPQPQLPYPQPQPF 6.17E+05 19 PYPQPQLPYPQPQPF 5.69E+05 22 QPQLPYPQPQPF 5.91E+05 24 QLPYPQPQPF 5.72E+05 26 PYPQPQPF 2.18E+05

1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 2.52E+08 1 LQLQPFPQ 3.76E+05 1 LQLQPFPQPQL 3.40E+05 1 LQLQPFPQPQLPY 4.46E+05 1 LQLQPFPQPQLPYPQ 9.54E+05 1 LQLQPFPQPQLPYPQPQLPY 9.56E+05 1 LQLQPFPQPQLPYPQPQLPYPQPQLPY 2.51E+06 1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQ 2.45E+06 L. johnsonii 7 PQPQLPYPQPQLPYPQPQLPYPQPQPF 2.24E+06 GDP6 11 LPYPQPQLPYPQPQLPYPQPQPF 6.62E+05 13 YPQPQLPYPQPQLPYPQPQPF 1.41E+06 14 PQPQLPYPQPQLPYPQPQPF 5.90E+05 17 QLPYPQPQLPYPQPQPF 4.83E+05 19 PYPQPQLPYPQPQPF 7.12E+05 20 YPQPQLPYPQPQPF 5.49E+05 21 PQPQLPYPQPQPF 2.51E+05 22 QPQLPYPQPQPF 4.95E+05 26 PYPQPQPF 2.84E+05

1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 2.42E+08 1 LQLQPFPQ 3.47E+05 1 LQLQPFPQPQL 3.58E+05 1 LQLQPFPQPQLPYPQ 9.46E+05 1 LQLQPFPQPQLPY 4.57E+05 1 LQLQPFPQPQLPYPQPQLPY 9.03E+05 1 LQLQPFPQPQLPYPQPQLPYPQPQLPY 2.33E+06 7 PQPQLPYPQPQLPYPQPQLPYPQPQPF 2.02E+06 8 QPQLPYPQPQLPYPQPQLPYPQPQPF 8.49E+05 L. ruminis 11 LPYPQPQLPYPQPQLPYPQPQPF 6.55E+05 GDP9 12 PYPQPQLPYPQPQLPYPQPQPF 1.21E+06 13 YPQPQLPYPQPQLPYPQPQPF 1.68E+06 19 PYPQPQLPYPQPQPF 9.05E+05 20 YPQPQLPYPQPQPF 1.08E+06 21 PQPQLPYPQPQPF 3.27E+05 22 QPQLPYPQPQPF 2.63E+05 24 QLPYPQPQPF 2.58E+05 25 LPYPQPQPF 1.57E+05 26 PYPQPQPF 3.69E+05 27 YPQPQPF 3.65E+05

1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 6.92E+06 4 QPFPQPQLPYPQPQLPY 1.01E+06 6 FPQPQLPYPQPQLPYPQPQLPYPQPQPF 3.84E+06 7 PQPQLPYPQPQLPYPQPQLPYPQPQPF 1.39E+07 8 QPQLPYPQPQLPYPQPQLPYPQPQPF 1.39E+06 12 PYPQPQLPYPQPQLPYPQPQPF 1.54E+06 13 YPQPQLPYPQPQLPYPQPQPF 1.51E+06 17 QLPYPQPQLPYPQPQPF 1.01E+06 L. salivarius 18 LPYPQPQLPYPQPQPF 5.57E+05 GDP12 19 PYPQPQLPYPQPQPF 1.05E+06 20 YPQPQLPYPQPQPF 1.16E+06 21 PQPQLPYPQPQPF 3.87E+05 22 QPQLPYPQPQPF 4.59E+05 23 PQLPYPQPQPF 3.27E+05 24 QLPYPQPQPF 2.87E+05 25 LPYPQPQPF 2.95E+05 27 YPQPQPF 3.92E+05

1 LQLQPFPQPQLPYPQPQLPYPQPQLPYPQPQPF 9.67E+06 4 QPFPQPQLPYPQPQLPYPQ 2.80E+05 4 QPFPQPQLPYPQPQLPYPQPQLPYPQ 2.64E+06 7 PQPQLPYPQPQLPYPQPQLPYPQPQPF 8.88E+06 GDP mix 8 QPQLPYPQPQLPYPQPQLPYPQPQPF 2.64E+06 12 PYPQPQLPYPQPQLPYPQPQPF 1.23E+06 13 YPQPQLPYPQPQLPYPQPQPF 1.18E+06 19 PYPQPQLPYPQPQPF 6.67E+05 20 YPQPQLPYPQPQPF 7.14E+05

170

Fig. S2. 16-mer degradation by the selected isolates. Bars represent the percent degradation quantitated by (black) RIDASCREEN® ELISA, (white) GlutenTox® ELISA, and (grey) and UHPLC-MS/MS-MRM. Data is presented as the average (n=2 ± STD) Absence of bar indicates no degradation detected.

Fig. S3 17-mer degradation by the selected isolates. Bars represent the percent degradation quantitated by (black) RIDASCREEN® ELISA, (white) GlutenTox® ELISA, and (grey) and UHPLC-MS/MS-MRM. Data is presented as the average (n=2 ± STD) Absence of bar indicates no degradation detected.

171

Fig S4 33-mer degradation by the selected isolates. Bars represent the percent degradation quantitated by (black) RIDASCREEN® ELISA, (white) GlutenTox® ELISA, and (grey) and UHPLC-MS/MS-MRM. Data is presented as the average (n=2 ± STD) Absence of bar indicates no degradation detected.

172

5.6 References

Alvarez-Sieiro P, Martin MC, Redruello B et al. Generation of food-grade recombinant

Lactobacillus casei delivering Myxococcus xanthus prolyl endopeptidase. Appl Microbiol

Biotechnol 2014;98:6689–700.

De Angelis M, Cassone A, Rizzello CG et al. Mechanism of degradation of immunogenic gluten epitopes from Triticum turgidum L. var. durum by sourdough lactobacilli and fungal proteases. Appl Environ Microbiol 2010;76:508–18.

De Angelis M, Rizzello CG, Fasano A et al. VSL#3 probiotic preparation has the capacity to hydrolyze gliadin polypeptides responsible for Celiac Sprue. Biochim Biophys Acta

2006;1762:80–93.

Arentz-Hansen H, Körner R, Molberg O et al. The intestinal T cell response to alpha-gliadin in adult celiac disease is focused on a single deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000;191:603–12.

Berntsson RP-A, Doeven MK, Fusetti F et al. The structural basis for peptide selection by the transport receptor OppA. EMBO J 2009;28:1332–40.

Bethune MT, Khosla C. Oral enzyme therapy for celiac sprue. Methods Enzymol

2012;502:241–71.

Caminero A, Herrán AR, Nistal E et al. Diversity of the cultivable human gut microbiome involved in gluten metabolism: isolation of microorganisms with potential interest for coeliac disease. FEMS Microbiol Ecol 2014;88:309–19.

Castillo M, Skene G, Roca M et al. Application of 16S rRNA gene-targetted fluorescence in situ hybridization and restriction fragment length polymorphism to study porcine microbiota along the gastrointestinal tract in response to different sources of dietary fibre. FEMS Microbiol Ecol

2007;59:138–46.

Christensen JE, Dudley EG, Pederson J a et al. Peptidases and amino acid catabolism in lactic acid bacteria. Antonie Van Leeuwenhoek 1999;76:217–46.

173

Daniel C, Roussel Y, Kleerebezem M et al. Recombinant lactic acid bacteria as mucosal biotherapeutic agents. Trends Biotechnol 2011;29:499–508.

Davila A-M, Blachier F, Gotteland M et al. Intestinal luminal nitrogen metabolism: role of the gut microbiota and consequences for the host. Pharmacol Res 2013;68:95–107.

Elli M, Zink R, Rytz A et al. Iron requirement of Lactobacillus spp. in completely chemically defined growth media. J Appl Microbiol 2000;88:695–703.

Fernandez-Feo M, Wei G, Blumenkranz G et al. The cultivable human oral gluten-degrading microbiome and its potential implications in coeliac disease and gluten sensitivity. Clin Microbiol

Infect 2013;19:E386-94.

Fernández M, Zúñiga M. Amino acid catabolic pathways of lactic acid bacteria. Crit Rev

Microbiol 2006;32:155–83.

Gass J, Bethune MT, Siegel M et al. Combination enzyme therapy for gastric digestion of dietary gluten in patients with celiac sprue. Gastroenterology 2007;133:472–80.

Gobbetti M, Giuseppe Rizzello C, Di Cagno R et al. Sourdough lactobacilli and celiac disease.

Food Microbiol 2007;24:187–96.

Green PHR, Cellier C. Celiac disease. N Engl J Med 2007;357:1731–43.

Heinritz SN, Mosenthin R, Weiss E. Use of pigs as a potential model for research into dietary modulation of the human gut microbiota. Nutr Res Rev 2013;26:191–209.

Jacobsen CN, Rosenfeldt Nielsen V, Hayford AE et al. Screening of probiotic activities of forty- seven strains of Lactobacillus spp. by in vitro techniques and evaluation of the colonization ability of five selected strains in humans. Appl Envir Microbiol 1999;65:4949–56.

Kaukinen K, Lindfors K, Mäki M. Advances in the treatment of coeliac disease: an immunopathogenic perspective. Nat Rev Gastroenterol Hepatol 2014;11:36–44.

Lähdeaho M-L, Kaukinen K, Laurila K et al. Glutenase ALV003 attenuates gluten-induced mucosal injury in patients with celiac disease. Gastroenterology 2014;146:1649–58.

174

Laparra JM, Sanz Y. Bifidobacteria inhibit the inflammatory response induced by gliadins in intestinal epithelial cells via modifications of toxic peptide generation during digestion. J Cell

Biochem 2010;109:801–7.

Leser TD, Amenuvor JZ, Jensen TK et al. Culture-independent analysis of gut bacteria: the pig gastrointestinal tract microbiota revisited. Appl Environ Microbiol 2002;68:673–90.

Lindfors K, Blomqvist T, Juuti-Uusitalo K et al. Live probiotic Bifidobacterium lactis bacteria inhibit the toxic effects induced by wheat gliadin in epithelial cell culture. Clin Exp Immunol

2008;152:552–8.

Liu M, Bayjanov JR, Renckens B et al. The proteolytic system of lactic acid bacteria revisited: a genomic comparison. BMC Genomics 2010;11:36.

Louis P, Scott KPP, Duncan SHH et al. Understanding the effects of diet on bacterial metabolism in the large intestine. J Appl Microbiol 2007;102:1197–208.

Ludvigsson JF, Rubio-Tapia A, van Dyke CT et al. Increasing incidence of celiac disease in a north american population. Am J Gastroenterol 2013;108:818–24.

Makarova K, Slesarev a, Wolf Y et al. Comparative genomics of the lactic acid bacteria. Proc

Natl Acad Sci U S A 2006;103:15611–6.

Martínez I, Wallace G, Zhang C et al. Diet-induced metabolic improvements in a hamster model of hypercholesterolemia are strongly linked to alterations of the gut microbiota. Appl

Environ Microbiol 2009;75:4175–84.

Meroth CB, Walter J, Hertel C et al. Monitoring the bacterial population dynamics in sourdough fermentation processes by using PCR-Denaturing Gradient Gel Electrophoresis. Appl Environ

Microbiol 2003;69:475–82.

Mitea C, Havenaar R, Drijfhout JW et al. Efficient degradation of gluten by a prolyl endoprotease in a gastrointestinal model: implications for coeliac disease. Gut 2008;57:25–32.

Nadal I, Donat E, Donant E et al. Imbalance in the composition of the duodenal microbiota of children with coeliac disease. J Med Microbiol 2007;56:1669–74.

175

Neish AS. Microbes in gastrointestinal health and disease. Gastroenterology 2009;136:65–

80.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Piper JL, Gray GM, Khosla C. Effect of prolyl endopeptidase on digestive-resistant gliadin peptides in vivo. J Pharmacol Exp Ther 2004;311:213–9.

Pyle GG, Paaso B, Anderson BE et al. Effect of pretreatment of food gluten with prolyl endopeptidase on gluten-induced malabsorption in celiac sprue. Clin Gastroenterol Hepatol

2005;3:687–94.

Rizzello CG, De Angelis M, Di Cagno R et al. Highly efficient gluten degradation by lactobacilli and fungal proteases during food processing: new perspectives for celiac disease. Appl Environ

Microbiol 2007;73:4499–507.

Rollán G, De Angelis M, Gobbetti M et al. Proteolytic activity and reduction of gliadin-like fractions by sourdough lactobacilli. J Appl Microbiol 2005;99:1495–502.

Di Sabatino A, Corazza GR. Coeliac disease. Lancet 2009;373:1480–93.

Sanz Y, Sánchez E, Marzotto M et al. Differences in faecal bacterial communities in coeliac and healthy children as detected by PCR and denaturing gradient gel electrophoresis. FEMS

Immunol Med Microbiol 2007;51:562–8.

Shan L, Marti T, Sollid LM et al. Comparative biochemical analysis of three bacterial prolyl endopeptidases: implications for coeliac sprue. Biochem J 2004;383:311–8.

Shan L, Molberg Ø, Parrot I et al. Structural basis for gluten intolerance in celiac sprue.

Science 2002;297:2275–9.

Shan L, Qiao S, Arentz-hansen H et al. Identification and analysis of multivalent proteolytically resistant peptides from gluten : Implications for Celiac Sprue. J Pro 2005;4:1732–41.

Siegel M, Bethune MT, Gass J et al. Rational design of combination enzyme therapy for celiac sprue. Chem Biol 2006;13:649–58.

176

Siegel M, Garber ME, Spencer AG et al. Safety, tolerability, and activity of ALV003: results from two phase 1 single, escalating-dose clinical trials. Dig Dis Sci 2012;57:440–50.

Sollid LM. Coeliac disease: dissecting a complex inflammatory disorder. Nat Rev Immunol

2002;2:647–55.

Sollid LM, Khosla C. Future therapeutic options for celiac disease. Nat Clin Pract

Gastroenterol Hepatol 2005;2:140–7.

Sollid LM, Qiao S-W, Anderson RP et al. Nomenclature and listing of celiac disease relevant gluten T-cell epitopes restricted by HLA-DQ molecules. Immunogenetics 2012;64:455–60.

de Sousa Moraes LF, Grzeskowiak LM, de Sales Teixeira TF et al. Intestinal microbiota and probiotics in celiac disease. Clin Microbiol Rev 2014;27:482–9.

Stepniak D, Spaenij-Dekking L, Mitea C et al. Highly efficient gluten degradation with a newly identified prolyl endoprotease: implications for celiac disease. Am J Physiol Gastrointest Liver

Physiol 2006;291:G621-9.

Tack GJ, van de Water JMW, Bruins MJ et al. Consumption of gluten with gluten-degrading enzyme by celiac patients: a pilot-study. World J Gastroenterol 2013;19:5837–47.

Tjellström B, Stenhammar L, Högberg L et al. Gut microflora associated characteristics in children with celiac disease. Am J Gastroenterol 2005;100:2784–8.

Tye-Din JA, Anderson RP, Ffrench RA et al. The effects of ALV003 pre-digestion of gluten on immune response and symptoms in celiac disease in vivo. Clin Immunol 2010a;134:289–95.

Tye-Din JA, Stewart JA, Dromey JA et al. Comprehensive, quantitative mapping of T cell epitopes in gluten in celiac disease. Sci Transl Med 2010b;2:41ra51.

Walter J. Ecological role of lactobacilli in the gastrointestinal tract: implications for fundamental and biomedical research. Appl Environ Microbiol 2008;74:4985–96.

Wells JM, Mercenier A. Mucosal delivery of therapeutic and prophylactic molecules using lactic acid bacteria. Nat Rev Microbiol 2008;6:349–62.

177

Zamakhchari M, Wei G, Dewhirst F et al. Identification of Rothia bacteria as gluten-degrading natural colonizers of the upper gastro-intestinal tract. PLoS One 2011;6:e24455.

178

6. Chapter six: Conclusions, future directions and implications

6.1 Conclusions and future directions

The phylogenomic framework presented in Chapter one allowed inferences regarding the natural history and lifestyle of lactobacilli and proposes a model for the evolution of distinct lifestyles within the genus, which range from free-living to strictly host-adapted and symbiotic species. Lifestyles of lactobacilli relate to their phylogeny and members of the same phylogenetic group or clade typically share important metabolic traits as well as characteristic aspects of their lifestyle. Many lactobacilli remain free-living, or adopted a nomadic lifestyle to use animal hosts for dispersal. Different groups in the genus adapted complex lifestyles that allowed them to colonize diverse animal hosts, ranging from insects to humans, with a different degree of host- specificity (Fig 1.6).

Chapter two explored the consequences of the evolution of L. reuteri with different vertebrate species. Findings from this work established that L. reuteri evolved jointly not only with mice but also chicken, and that this process of co-evolution resulted in specialization. Those strains that evolved jointly with their hosts showed elevated fitness when compared with strains that share no natural history with that same host.

Work in Chapter four revealed that the ability of human strains to cross-feed from 1,2 propanediol (1,2 PD) could be relevant for their competitiveness in the gut. Rhamnose, one of the precursors of 1,2 PD reaches the gut through pectin and hemicellulosic polysaccharides present in plant fibers. Given that modern diets are significantly low in plant fibers (Deehan and Walter,

2016) a reduced amount of rhamnose, and consequentially of 1,2 PD, could mean that the conditions in which L. reuteri evolved in humans are no longer existent and would explain findings in Chapter two.

179

As reviewed in Chapter one, a common strategy by which lactobacilli maintain host-specificity is by adhering to the host’s tissue (Fig 1.4). Rodent strains of L. reuteri adhere to the mouse forestomach by forming biofilms, a mechanism that is restricted to rodent strains (Frese et al.

2013). Work in Chapter three characterized a two component system (TCS70529-30) which is highly conserved in rodent strains and had been hypothesized to have a regulatory role in biofilm formation (Frese et al. 2011). Findings from in vivo and in vitro experiments confirmed that

TCS70529-30 has a role in biofilm formation, most likely by regulating genes involved in the development of the exopolymeric matrix. Deletion of the individual genes of the two component system resulted in highly disparate phenotypes, with histidine kinase mutants producing robust biofilms, while response regulator mutants produced fragile biofilms (Fig 3.2-3.7). Future studies are thus required to fully elucidate the role of TCS70529-30 in biofilm formation. Efforts to determine the transcriptional targets of TCS70529-30 are currently underway.

One powerful characteristics of L. reuteri as model species to study adaptation to the vertebrate gut, is that strains from different hosts differ at the genetic level. These differences often correlate with the conditions encountered in the gut of each corresponding host (Frese et al. 2011). For example, contrary to rodents strains, human isolates conserve a 58-gene cluster encoding a 1,2 propanediol dehydratase PduCDE. Chapter four explored the role of PduCDE enzyme in the ecological performance of L. reuteri. From a series of in vitro cross-feeding experiments it was concluded that L. reuteri can cross-feed from 1,2 PD-produced by strains of

Escherichia coli and Bifidobacterium breve from the fermentation of deoxyhexoses found in the gut. During fermentation of hexoses, 1,2 PD serves as an electron acceptor increasing the metabolic efficiency of L. reuteri, a factor that could be pivotal to the competitiveness of L. reuteri in the gut ecosystems. These findings are exciting because L-rhamnose, a precursor of 1,2 PD, can be introduced into the human gut via dietary plant-derived polysaccharides. Based on these results, our group is currently planning dietary intervention trials aimed to determine if increasing

180

the consumption of rhamnose-containing foods, and thus the availability of 1,2 PD in the gut, can support the persistence of L. reuteri in humans.

Finally, work in Chapter five (Duar et al. 2015) explored an innovative strategy to select and exploit the metabolic capacity of lactobacilli adapted to the vertebrate gut. Strains of L. salivarius,

L. amylovorus. L ruminis and L. johnsonii isolated from the pig’s gastrointestinal tract demonstrated the ability to degrade metastable peptides implicated in the pathology of celiac disease. Although these strains are not of human origin, they are likely to be active the conditions of the proximal gastrointestinal tract. This characterictic could be advantageous as gluten peptides would need to be degraded before reaching the intestinal epithelium. Furthermore, findings in Chapter 2 demonstrated that porcine strains of L. reuteri are competitive in the human gut. The gastrointestinal tract of pigs are anatomically and physiologically similar (Heinritz,

Mosenthin and Weiss 2013). However, contrary to pigs, humans harbor relatively low amounts of lactobacilli. Therefore, sourcing strains from the porcine hosts could pose a suitable strategy to select for strains with probiotic potential in humans. Together these chapters provide a basis for the selection of lactobacilli strains aimed to be competitive and physiologically active in the human gastrointestinal tract.

6.2 Implications

The implications of findings from this dissertation extend beyond its contribution to a basic understanding of the biology and ecology of the genus Lactobacillus. Humans have essentially

“domesticated” lactobacilli for use in food and feed production, an increased understanding of the origin of these microbes and their function in nature will therefore facilitate the selection of strains for such applications. Attributes that lactobacilli that evolved in their natural habitats, such as metabolic functions, antagonism towards other members of microbial communities, and their impact on host species, can be exploited once understood. In addition, an understanding of host- associated lactobacilli might allow the development of strategies to support their populations or

181

beneficial metabolic activities through dietary intervention. For example, strong attention has been paid towards the unsolved decline in the population of honey bees (Goulson et al. 2015; Engel et al. 2016). Their paramount importance as pollinators of agricultural crops justifies efforts to understand and better manage their symbiotic interactions with microbes as a tool to preserve insect health (Engel et al. 2016). Similar considerations also apply to humans and farm animals, many of which maintain dominant population of lactobacilli among their microbiota (i.e. swine and poultry). The framework presented in Chapters one, two and five lay a foundation of strategies to select strains for a whole range of biotechnological and therapeutic applications.

Furthermore, as demonstrated in Chapter two and by others (Oh et al. 2010; Frese et al.

2011), host-adapted strains of lactobacilli show a higher ecological fitness in their respective hosts

Therefore, host-specific lactobacilli are likely to be more competitive when administered as a probiotics, compared to strains that do not share an evolutionary history with the host. Higher fitness is relevant for the development of probiotics supposed to outcompete pathogens, and it is likely to be associated with higher metabolic activity in the host niche, which could lead to an increased production of metabolic compounds that define probiotic activity. In addition, stable transmission of bacterial symbionts over evolutionary times promotes traits that enhance partner performance (Herre et al. 1999; Sachs et al. 2004; Douglas 2008). Providing this theory holds true for the relationship between lactobacilli and animal hosts, then host-adapted Lactobacillus strains that share an evolutionary fate with their host are more likely to possess adaptive traits that enhance health of their host.

Such evolutionary aspects have rarely been considered for the selection of strains for specific applications. It is a logical working hypothesis that host-adapted Lactobacillus strains will show higher levels of ecological performance when used as probiotics, possess beneficial traits that enhance host fitness, and are likely to establish interactions with the host immune system that are characterized by tolerance (Walter, Britton and Roos 2011). Conversely, if the aim is to stimulate the immune system, selection of species or strains that lack a joint evolution with a host

182

may be a more sensible approach. This proved true for L. reuteri in which strains from the autochthonous human lineage had an anti-inflammatory effect in human myeloid cells while strains associated with the phylogenetic lineage that evolved with poultry had a rather stimulatory immune effect (Spinler et al. 2014). These findings highlight the functional significance of the natural history of lactobacilli for probiotic functions. Although one cannot generalize what constitutes a better probiotic - host-adapted or not – the evolutionary history of a strain will fundamentally influence its functionality. A consideration of the natural history of lactobacilli will therefore aid in the more systematic and targeted selection of optimal strains for specific therapeutic applications. The framework established in Chapters one and two can clearly provide guidance for such a selection.

Finally, if as hypothesized, findings in Chapter two regarding the inability of human strains of

L. reuteri to persist in the human gut, are explained by findings in Chapter four, then increasing rhamnose in the diet could represents a suitable strategy to increase production of 1,2 PD in order to support the persistence of L. reuteri in humans. Dietary interventions aimed to increase rhamose in the diet are therefore warranted to shed light into the relevance of 1,2 PD in vivo and would provide a basis for dietary strategies aimed to support other human gut symbionts.

6.3 References

Bouckaert R, Heled J, Kühnert D et al. BEAST 2: A Software Platform for Bayesian

Evolutionary Analysis. PLoS Comput Biol 2014;10:e1003537.

Deehan EC, Walter J. The fiber gap and the disappearing gut microbiome: implications for human nutrition. Trends Endocrinol Metab 2016;27:239–42.

Douglas AE. Conflict, cheats and the persistence of symbioses. New Phytol 2008;177:849–

58.

183

Duar RM, Clark KJ, Patil PB et al. Identification and characterization of intestinal lactobacilli strains capable of degrading immunotoxic peptides present in gluten. J Appl Microbiol 2015, DOI:

10.1111/jam.12687. (Chapter five of this thesis)

Engel P, Kwong WK, McFrederick Q et al. The Bee Microbiome: Impact on bee health and model for evolution and ecology of host-microbe interactions. MBio 2016;7:e02164-15.

Frese SA, Benson AK, Tannock GW et al. The evolution of host specialization in the vertebrate gut symbiont Lactobacillus reuteri. PLoS Genet 2011;7:16.

Goulson D, Nicholls E, Botias C et al. Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science 2015;347:1255957–1255957.

Heinritz SN, Mosenthin R, Weiss E. Use of pigs as a potential model for research into dietary modulation of the human gut microbiota. Nutr Res Rev 2013;26:191–209.

Herre E, Knowlton N, Mueller U et al. The evolution of mutualisms: exploring the paths between conflict and cooperation. Trends Ecol Evol 1999;14:49–53.

Maddison W, Maddisson D. Mesquite: A modular system for evolutionary analysis. Version

3.2. 2017.

Oh PL, Benson AK, Peterson DA et al. Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J 2010;4:377–87.

Orla-Jensen S. The Lactic Acid Bacteria. Copenhagen: Andr. Fred. Host & Son, 1919.

Sachs JL, Mueller UG, Wilcox TP et al. The evolution of cooperation. Q Rev Biol 2004;79:135–

60.

Spinler JK, Sontakke A, Hollister EB et al. From prediction to function using evolutionary genomics: human-specific ecotypes of Lactobacillus reuteri have diverse probiotic functions.

Genome Biol Evol 2014;6:1772–89.

Walter J, Britton RA, Roos S. Host-microbial symbiosis in the vertebrate gastrointestinal tract and the Lactobacillus reuteri paradigm. Proc Natl Acad Sci U S A 2011;108 Suppl:4645–52.

184

185

REFERENCES

Chapter one

Ahrné S, Nobaek S, Jeppsson B et al. The normal Lactobacillus flora of healthy human rectal and oral mucosa. J Appl Microbiol 1998;85:88–94.

Alcántara C, Zúñiga M. Proteomic and transcriptomic analysis of the response to bile stress of Lactobacillus casei BL23. Microbiology 2012;158:1206–18.

Anderson AC, Sanunu M, Schneider C et al. Rapid species-level identification of vaginal and oral lactobacilli using MALDI-TOF MS analysis and 16S rDNA sequencing. BMC Microbiol

2014;14:312.

Anderson KEK, Sheehan THTTH, Mott BMB et al. Microbial Ecology of the Hive and

Pollination Landscape: Bacterial Associates from Floral Nectar, the Alimentary Tract and Stored

Food of Honey Bees (Apis mellifera). Gerardo NM (ed.). PLoS One 2013;8:e83125.

Andreevskaya M. Ecological fitness and interspecies interactions of food-spoilage-associated lactic acid bacteria: insights from the genome analyses and transcriptome profiles. 2017.

Battistuzzi FU, Feijao A, Hedges SB et al. A genomic timescale of prokaryote evolution: insights into the origin of methanogenesis, phototrophy, and the colonization of land. BMC Evol

Biol 2004;4:44.

van Bokhorst-van de Veen H, Smelt MJ, Wels M et al. Genotypic adaptations associated with prolonged persistence of Lactobacillus plantarum in the murine digestive tract. Biotechnol J

2013;8:895–904.

Bringel F, Curk MC, Hubert JC. Characterization of lactobacilli by southern-type hybridization with a Lactobacillus plantarum pyrDFE probe. Int J Syst Bacteriol 1996;46:588–94.

186